blob: 1164015784011bad2a20fe14a9eab846c013153a [file] [log] [blame]
rjw1f884582022-01-06 17:20:42 +08001/*
2 * ext4_jbd2.h
3 *
4 * Written by Stephen C. Tweedie <sct@redhat.com>, 1999
5 *
6 * Copyright 1998--1999 Red Hat corp --- All Rights Reserved
7 *
8 * This file is part of the Linux kernel and is made available under
9 * the terms of the GNU General Public License, version 2, or at your
10 * option, any later version, incorporated herein by reference.
11 *
12 * Ext4-specific journaling extensions.
13 */
14
15#ifndef _EXT4_JBD2_H
16#define _EXT4_JBD2_H
17
18#include <linux/fs.h>
19#include <linux/jbd2.h>
20#include "ext4.h"
21
22#define EXT4_JOURNAL(inode) (EXT4_SB((inode)->i_sb)->s_journal)
23
24/* Define the number of blocks we need to account to a transaction to
25 * modify one block of data.
26 *
27 * We may have to touch one inode, one bitmap buffer, up to three
28 * indirection blocks, the group and superblock summaries, and the data
29 * block to complete the transaction.
30 *
31 * For extents-enabled fs we may have to allocate and modify up to
32 * 5 levels of tree, data block (for each of these we need bitmap + group
33 * summaries), root which is stored in the inode, sb
34 */
35
36#define EXT4_SINGLEDATA_TRANS_BLOCKS(sb) \
37 (ext4_has_feature_extents(sb) ? 20U : 8U)
38
39/* Extended attribute operations touch at most two data buffers,
40 * two bitmap buffers, and two group summaries, in addition to the inode
41 * and the superblock, which are already accounted for. */
42
43#define EXT4_XATTR_TRANS_BLOCKS 6U
44
45/* Define the minimum size for a transaction which modifies data. This
46 * needs to take into account the fact that we may end up modifying two
47 * quota files too (one for the group, one for the user quota). The
48 * superblock only gets updated once, of course, so don't bother
49 * counting that again for the quota updates. */
50
51#define EXT4_DATA_TRANS_BLOCKS(sb) (EXT4_SINGLEDATA_TRANS_BLOCKS(sb) + \
52 EXT4_XATTR_TRANS_BLOCKS - 2 + \
53 EXT4_MAXQUOTAS_TRANS_BLOCKS(sb))
54
55/*
56 * Define the number of metadata blocks we need to account to modify data.
57 *
58 * This include super block, inode block, quota blocks and xattr blocks
59 */
60#define EXT4_META_TRANS_BLOCKS(sb) (EXT4_XATTR_TRANS_BLOCKS + \
61 EXT4_MAXQUOTAS_TRANS_BLOCKS(sb))
62
63/* Define an arbitrary limit for the amount of data we will anticipate
64 * writing to any given transaction. For unbounded transactions such as
65 * write(2) and truncate(2) we can write more than this, but we always
66 * start off at the maximum transaction size and grow the transaction
67 * optimistically as we go. */
68
69#define EXT4_MAX_TRANS_DATA 64U
70
71/* We break up a large truncate or write transaction once the handle's
72 * buffer credits gets this low, we need either to extend the
73 * transaction or to start a new one. Reserve enough space here for
74 * inode, bitmap, superblock, group and indirection updates for at least
75 * one block, plus two quota updates. Quota allocations are not
76 * needed. */
77
78#define EXT4_RESERVE_TRANS_BLOCKS 12U
79
80/*
81 * Number of credits needed if we need to insert an entry into a
82 * directory. For each new index block, we need 4 blocks (old index
83 * block, new index block, bitmap block, bg summary). For normal
84 * htree directories there are 2 levels; if the largedir feature
85 * enabled it's 3 levels.
86 */
87#define EXT4_INDEX_EXTRA_TRANS_BLOCKS 12U
88
89#ifdef CONFIG_QUOTA
90/* Amount of blocks needed for quota update - we know that the structure was
91 * allocated so we need to update only data block */
92#define EXT4_QUOTA_TRANS_BLOCKS(sb) ((test_opt(sb, QUOTA) ||\
93 ext4_has_feature_quota(sb)) ? 1 : 0)
94/* Amount of blocks needed for quota insert/delete - we do some block writes
95 * but inode, sb and group updates are done only once */
96#define EXT4_QUOTA_INIT_BLOCKS(sb) ((test_opt(sb, QUOTA) ||\
97 ext4_has_feature_quota(sb)) ?\
98 (DQUOT_INIT_ALLOC*(EXT4_SINGLEDATA_TRANS_BLOCKS(sb)-3)\
99 +3+DQUOT_INIT_REWRITE) : 0)
100
101#define EXT4_QUOTA_DEL_BLOCKS(sb) ((test_opt(sb, QUOTA) ||\
102 ext4_has_feature_quota(sb)) ?\
103 (DQUOT_DEL_ALLOC*(EXT4_SINGLEDATA_TRANS_BLOCKS(sb)-3)\
104 +3+DQUOT_DEL_REWRITE) : 0)
105#else
106#define EXT4_QUOTA_TRANS_BLOCKS(sb) 0
107#define EXT4_QUOTA_INIT_BLOCKS(sb) 0
108#define EXT4_QUOTA_DEL_BLOCKS(sb) 0
109#endif
110#define EXT4_MAXQUOTAS_TRANS_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_TRANS_BLOCKS(sb))
111#define EXT4_MAXQUOTAS_INIT_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_INIT_BLOCKS(sb))
112#define EXT4_MAXQUOTAS_DEL_BLOCKS(sb) (EXT4_MAXQUOTAS*EXT4_QUOTA_DEL_BLOCKS(sb))
113
114/*
115 * Ext4 handle operation types -- for logging purposes
116 */
117#define EXT4_HT_MISC 0
118#define EXT4_HT_INODE 1
119#define EXT4_HT_WRITE_PAGE 2
120#define EXT4_HT_MAP_BLOCKS 3
121#define EXT4_HT_DIR 4
122#define EXT4_HT_TRUNCATE 5
123#define EXT4_HT_QUOTA 6
124#define EXT4_HT_RESIZE 7
125#define EXT4_HT_MIGRATE 8
126#define EXT4_HT_MOVE_EXTENTS 9
127#define EXT4_HT_XATTR 10
128#define EXT4_HT_EXT_CONVERT 11
129#define EXT4_HT_MAX 12
130
131/**
132 * struct ext4_journal_cb_entry - Base structure for callback information.
133 *
134 * This struct is a 'seed' structure for a using with your own callback
135 * structs. If you are using callbacks you must allocate one of these
136 * or another struct of your own definition which has this struct
137 * as it's first element and pass it to ext4_journal_callback_add().
138 */
139struct ext4_journal_cb_entry {
140 /* list information for other callbacks attached to the same handle */
141 struct list_head jce_list;
142
143 /* Function to call with this callback structure */
144 void (*jce_func)(struct super_block *sb,
145 struct ext4_journal_cb_entry *jce, int error);
146
147 /* user data goes here */
148};
149
150/**
151 * ext4_journal_callback_add: add a function to call after transaction commit
152 * @handle: active journal transaction handle to register callback on
153 * @func: callback function to call after the transaction has committed:
154 * @sb: superblock of current filesystem for transaction
155 * @jce: returned journal callback data
156 * @rc: journal state at commit (0 = transaction committed properly)
157 * @jce: journal callback data (internal and function private data struct)
158 *
159 * The registered function will be called in the context of the journal thread
160 * after the transaction for which the handle was created has completed.
161 *
162 * No locks are held when the callback function is called, so it is safe to
163 * call blocking functions from within the callback, but the callback should
164 * not block or run for too long, or the filesystem will be blocked waiting for
165 * the next transaction to commit. No journaling functions can be used, or
166 * there is a risk of deadlock.
167 *
168 * There is no guaranteed calling order of multiple registered callbacks on
169 * the same transaction.
170 */
171static inline void _ext4_journal_callback_add(handle_t *handle,
172 struct ext4_journal_cb_entry *jce)
173{
174 /* Add the jce to transaction's private list */
175 list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
176}
177
178static inline void ext4_journal_callback_add(handle_t *handle,
179 void (*func)(struct super_block *sb,
180 struct ext4_journal_cb_entry *jce,
181 int rc),
182 struct ext4_journal_cb_entry *jce)
183{
184 struct ext4_sb_info *sbi =
185 EXT4_SB(handle->h_transaction->t_journal->j_private);
186
187 /* Add the jce to transaction's private list */
188 jce->jce_func = func;
189 spin_lock(&sbi->s_md_lock);
190 _ext4_journal_callback_add(handle, jce);
191 spin_unlock(&sbi->s_md_lock);
192}
193
194
195/**
196 * ext4_journal_callback_del: delete a registered callback
197 * @handle: active journal transaction handle on which callback was registered
198 * @jce: registered journal callback entry to unregister
199 * Return true if object was successfully removed
200 */
201static inline bool ext4_journal_callback_try_del(handle_t *handle,
202 struct ext4_journal_cb_entry *jce)
203{
204 bool deleted;
205 struct ext4_sb_info *sbi =
206 EXT4_SB(handle->h_transaction->t_journal->j_private);
207
208 spin_lock(&sbi->s_md_lock);
209 deleted = !list_empty(&jce->jce_list);
210 list_del_init(&jce->jce_list);
211 spin_unlock(&sbi->s_md_lock);
212 return deleted;
213}
214
215int
216ext4_mark_iloc_dirty(handle_t *handle,
217 struct inode *inode,
218 struct ext4_iloc *iloc);
219
220/*
221 * On success, We end up with an outstanding reference count against
222 * iloc->bh. This _must_ be cleaned up later.
223 */
224
225int ext4_reserve_inode_write(handle_t *handle, struct inode *inode,
226 struct ext4_iloc *iloc);
227
228int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode);
229
230int ext4_expand_extra_isize(struct inode *inode,
231 unsigned int new_extra_isize,
232 struct ext4_iloc *iloc);
233/*
234 * Wrapper functions with which ext4 calls into JBD.
235 */
236int __ext4_journal_get_write_access(const char *where, unsigned int line,
237 handle_t *handle, struct buffer_head *bh);
238
239int __ext4_forget(const char *where, unsigned int line, handle_t *handle,
240 int is_metadata, struct inode *inode,
241 struct buffer_head *bh, ext4_fsblk_t blocknr);
242
243int __ext4_journal_get_create_access(const char *where, unsigned int line,
244 handle_t *handle, struct buffer_head *bh);
245
246int __ext4_handle_dirty_metadata(const char *where, unsigned int line,
247 handle_t *handle, struct inode *inode,
248 struct buffer_head *bh);
249
250int __ext4_handle_dirty_super(const char *where, unsigned int line,
251 handle_t *handle, struct super_block *sb);
252
253#define ext4_journal_get_write_access(handle, bh) \
254 __ext4_journal_get_write_access(__func__, __LINE__, (handle), (bh))
255#define ext4_forget(handle, is_metadata, inode, bh, block_nr) \
256 __ext4_forget(__func__, __LINE__, (handle), (is_metadata), (inode), \
257 (bh), (block_nr))
258#define ext4_journal_get_create_access(handle, bh) \
259 __ext4_journal_get_create_access(__func__, __LINE__, (handle), (bh))
260#define ext4_handle_dirty_metadata(handle, inode, bh) \
261 __ext4_handle_dirty_metadata(__func__, __LINE__, (handle), (inode), \
262 (bh))
263#define ext4_handle_dirty_super(handle, sb) \
264 __ext4_handle_dirty_super(__func__, __LINE__, (handle), (sb))
265
266handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line,
267 int type, int blocks, int rsv_blocks);
268int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle);
269
270#define EXT4_NOJOURNAL_MAX_REF_COUNT ((unsigned long) 4096)
271
272/* Note: Do not use this for NULL handles. This is only to determine if
273 * a properly allocated handle is using a journal or not. */
274static inline int ext4_handle_valid(handle_t *handle)
275{
276 if ((unsigned long)handle < EXT4_NOJOURNAL_MAX_REF_COUNT)
277 return 0;
278 return 1;
279}
280
281static inline void ext4_handle_sync(handle_t *handle)
282{
283 if (ext4_handle_valid(handle))
284 handle->h_sync = 1;
285}
286
287static inline int ext4_handle_is_aborted(handle_t *handle)
288{
289 if (ext4_handle_valid(handle))
290 return is_handle_aborted(handle);
291 return 0;
292}
293
294static inline int ext4_handle_has_enough_credits(handle_t *handle, int needed)
295{
296 if (ext4_handle_valid(handle) && handle->h_buffer_credits < needed)
297 return 0;
298 return 1;
299}
300
301#define ext4_journal_start_sb(sb, type, nblocks) \
302 __ext4_journal_start_sb((sb), __LINE__, (type), (nblocks), 0)
303
304#define ext4_journal_start(inode, type, nblocks) \
305 __ext4_journal_start((inode), __LINE__, (type), (nblocks), 0)
306
307#define ext4_journal_start_with_reserve(inode, type, blocks, rsv_blocks) \
308 __ext4_journal_start((inode), __LINE__, (type), (blocks), (rsv_blocks))
309
310static inline handle_t *__ext4_journal_start(struct inode *inode,
311 unsigned int line, int type,
312 int blocks, int rsv_blocks)
313{
314 return __ext4_journal_start_sb(inode->i_sb, line, type, blocks,
315 rsv_blocks);
316}
317
318#define ext4_journal_stop(handle) \
319 __ext4_journal_stop(__func__, __LINE__, (handle))
320
321#define ext4_journal_start_reserved(handle, type) \
322 __ext4_journal_start_reserved((handle), __LINE__, (type))
323
324handle_t *__ext4_journal_start_reserved(handle_t *handle, unsigned int line,
325 int type);
326
327static inline void ext4_journal_free_reserved(handle_t *handle)
328{
329 if (ext4_handle_valid(handle))
330 jbd2_journal_free_reserved(handle);
331}
332
333static inline handle_t *ext4_journal_current_handle(void)
334{
335 return journal_current_handle();
336}
337
338static inline int ext4_journal_extend(handle_t *handle, int nblocks)
339{
340 if (ext4_handle_valid(handle))
341 return jbd2_journal_extend(handle, nblocks);
342 return 0;
343}
344
345static inline int ext4_journal_restart(handle_t *handle, int nblocks)
346{
347 if (ext4_handle_valid(handle))
348 return jbd2_journal_restart(handle, nblocks);
349 return 0;
350}
351
352static inline int ext4_journal_blocks_per_page(struct inode *inode)
353{
354 if (EXT4_JOURNAL(inode) != NULL)
355 return jbd2_journal_blocks_per_page(inode);
356 return 0;
357}
358
359static inline int ext4_journal_force_commit(journal_t *journal)
360{
361 if (journal)
362 return jbd2_journal_force_commit(journal);
363 return 0;
364}
365
366static inline int ext4_jbd2_inode_add_write(handle_t *handle,
367 struct inode *inode, loff_t start_byte, loff_t length)
368{
369 if (ext4_handle_valid(handle))
370 return jbd2_journal_inode_ranged_write(handle,
371 EXT4_I(inode)->jinode, start_byte, length);
372 return 0;
373}
374
375static inline int ext4_jbd2_inode_add_wait(handle_t *handle,
376 struct inode *inode, loff_t start_byte, loff_t length)
377{
378 if (ext4_handle_valid(handle))
379 return jbd2_journal_inode_ranged_wait(handle,
380 EXT4_I(inode)->jinode, start_byte, length);
381 return 0;
382}
383
384static inline void ext4_update_inode_fsync_trans(handle_t *handle,
385 struct inode *inode,
386 int datasync)
387{
388 struct ext4_inode_info *ei = EXT4_I(inode);
389
390 if (ext4_handle_valid(handle) && !is_handle_aborted(handle)) {
391 ei->i_sync_tid = handle->h_transaction->t_tid;
392 if (datasync)
393 ei->i_datasync_tid = handle->h_transaction->t_tid;
394 }
395}
396
397/* super.c */
398int ext4_force_commit(struct super_block *sb);
399
400/*
401 * Ext4 inode journal modes
402 */
403#define EXT4_INODE_JOURNAL_DATA_MODE 0x01 /* journal data mode */
404#define EXT4_INODE_ORDERED_DATA_MODE 0x02 /* ordered data mode */
405#define EXT4_INODE_WRITEBACK_DATA_MODE 0x04 /* writeback data mode */
406
407static inline int ext4_inode_journal_mode(struct inode *inode)
408{
409 if (EXT4_JOURNAL(inode) == NULL)
410 return EXT4_INODE_WRITEBACK_DATA_MODE; /* writeback */
411 /* We do not support data journalling with delayed allocation */
412 if (!S_ISREG(inode->i_mode) ||
413 test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA ||
414 (ext4_test_inode_flag(inode, EXT4_INODE_JOURNAL_DATA) &&
415 !test_opt(inode->i_sb, DELALLOC))) {
416 /* We do not support data journalling for encrypted data */
417 if (S_ISREG(inode->i_mode) && ext4_encrypted_inode(inode))
418 return EXT4_INODE_ORDERED_DATA_MODE; /* ordered */
419 return EXT4_INODE_JOURNAL_DATA_MODE; /* journal data */
420 }
421 if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA)
422 return EXT4_INODE_ORDERED_DATA_MODE; /* ordered */
423 if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)
424 return EXT4_INODE_WRITEBACK_DATA_MODE; /* writeback */
425 BUG();
426}
427
428static inline int ext4_should_journal_data(struct inode *inode)
429{
430 return ext4_inode_journal_mode(inode) & EXT4_INODE_JOURNAL_DATA_MODE;
431}
432
433static inline int ext4_should_order_data(struct inode *inode)
434{
435 return ext4_inode_journal_mode(inode) & EXT4_INODE_ORDERED_DATA_MODE;
436}
437
438static inline int ext4_should_writeback_data(struct inode *inode)
439{
440 return ext4_inode_journal_mode(inode) & EXT4_INODE_WRITEBACK_DATA_MODE;
441}
442
443/*
444 * This function controls whether or not we should try to go down the
445 * dioread_nolock code paths, which makes it safe to avoid taking
446 * i_mutex for direct I/O reads. This only works for extent-based
447 * files, and it doesn't work if data journaling is enabled, since the
448 * dioread_nolock code uses b_private to pass information back to the
449 * I/O completion handler, and this conflicts with the jbd's use of
450 * b_private.
451 */
452static inline int ext4_should_dioread_nolock(struct inode *inode)
453{
454 if (!test_opt(inode->i_sb, DIOREAD_NOLOCK))
455 return 0;
456 if (!S_ISREG(inode->i_mode))
457 return 0;
458 if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
459 return 0;
460 if (ext4_should_journal_data(inode))
461 return 0;
462 return 1;
463}
464
465#endif /* _EXT4_JBD2_H */