journal.c 75.1 KB
Newer Older
1
/*
2
 * linux/fs/jbd2/journal.c
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
 *
 * Written by Stephen C. Tweedie <sct@redhat.com>, 1998
 *
 * Copyright 1998 Red Hat corp --- All Rights Reserved
 *
 * This file is part of the Linux kernel and is made available under
 * the terms of the GNU General Public License, version 2, or at your
 * option, any later version, incorporated herein by reference.
 *
 * Generic filesystem journal-writing code; part of the ext2fs
 * journaling system.
 *
 * This file manages journals: areas of disk reserved for logging
 * transactional updates.  This includes the kernel journaling thread
 * which is responsible for scheduling updates to the log.
 *
 * We do not actually manage the physical storage of the journal in this
 * file: that is left to a per-journal policy function, which allows us
 * to store the journal within a filesystem-specified area for ext2
 * journaling (ext2 can use a reserved inode for storing the log).
 */

#include <linux/module.h>
#include <linux/time.h>
#include <linux/fs.h>
28
#include <linux/jbd2.h>
29 30 31 32
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/mm.h>
33
#include <linux/freezer.h>
34 35 36 37
#include <linux/pagemap.h>
#include <linux/kthread.h>
#include <linux/poison.h>
#include <linux/proc_fs.h>
38
#include <linux/seq_file.h>
39
#include <linux/math64.h>
40
#include <linux/hash.h>
41 42
#include <linux/log2.h>
#include <linux/vmalloc.h>
43
#include <linux/backing-dev.h>
44
#include <linux/bitops.h>
45
#include <linux/ratelimit.h>
46 47 48

#define CREATE_TRACE_POINTS
#include <trace/events/jbd2.h>
49 50 51 52

#include <asm/uaccess.h>
#include <asm/page.h>

53 54 55 56 57 58 59 60
#ifdef CONFIG_JBD2_DEBUG
ushort jbd2_journal_enable_debug __read_mostly;
EXPORT_SYMBOL(jbd2_journal_enable_debug);

module_param_named(jbd2_debug, jbd2_journal_enable_debug, ushort, 0644);
MODULE_PARM_DESC(jbd2_debug, "Debugging level for jbd2");
#endif

61 62 63 64 65 66 67
EXPORT_SYMBOL(jbd2_journal_extend);
EXPORT_SYMBOL(jbd2_journal_stop);
EXPORT_SYMBOL(jbd2_journal_lock_updates);
EXPORT_SYMBOL(jbd2_journal_unlock_updates);
EXPORT_SYMBOL(jbd2_journal_get_write_access);
EXPORT_SYMBOL(jbd2_journal_get_create_access);
EXPORT_SYMBOL(jbd2_journal_get_undo_access);
J
Joel Becker 已提交
68
EXPORT_SYMBOL(jbd2_journal_set_triggers);
69 70
EXPORT_SYMBOL(jbd2_journal_dirty_metadata);
EXPORT_SYMBOL(jbd2_journal_forget);
71 72 73
#if 0
EXPORT_SYMBOL(journal_sync_buffer);
#endif
74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
EXPORT_SYMBOL(jbd2_journal_flush);
EXPORT_SYMBOL(jbd2_journal_revoke);

EXPORT_SYMBOL(jbd2_journal_init_dev);
EXPORT_SYMBOL(jbd2_journal_init_inode);
EXPORT_SYMBOL(jbd2_journal_check_used_features);
EXPORT_SYMBOL(jbd2_journal_check_available_features);
EXPORT_SYMBOL(jbd2_journal_set_features);
EXPORT_SYMBOL(jbd2_journal_load);
EXPORT_SYMBOL(jbd2_journal_destroy);
EXPORT_SYMBOL(jbd2_journal_abort);
EXPORT_SYMBOL(jbd2_journal_errno);
EXPORT_SYMBOL(jbd2_journal_ack_err);
EXPORT_SYMBOL(jbd2_journal_clear_err);
EXPORT_SYMBOL(jbd2_log_wait_commit);
89
EXPORT_SYMBOL(jbd2_log_start_commit);
90 91 92 93 94 95 96
EXPORT_SYMBOL(jbd2_journal_start_commit);
EXPORT_SYMBOL(jbd2_journal_force_commit_nested);
EXPORT_SYMBOL(jbd2_journal_wipe);
EXPORT_SYMBOL(jbd2_journal_blocks_per_page);
EXPORT_SYMBOL(jbd2_journal_invalidatepage);
EXPORT_SYMBOL(jbd2_journal_try_to_free_buffers);
EXPORT_SYMBOL(jbd2_journal_force_commit);
97 98
EXPORT_SYMBOL(jbd2_journal_inode_add_write);
EXPORT_SYMBOL(jbd2_journal_inode_add_wait);
99 100 101
EXPORT_SYMBOL(jbd2_journal_init_jbd_inode);
EXPORT_SYMBOL(jbd2_journal_release_jbd_inode);
EXPORT_SYMBOL(jbd2_journal_begin_ordered_truncate);
102
EXPORT_SYMBOL(jbd2_inode_cache);
103 104

static void __journal_abort_soft (journal_t *journal, int errno);
105
static int jbd2_journal_create_slab(size_t slab_size);
106

107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124
#ifdef CONFIG_JBD2_DEBUG
void __jbd2_debug(int level, const char *file, const char *func,
		  unsigned int line, const char *fmt, ...)
{
	struct va_format vaf;
	va_list args;

	if (level > jbd2_journal_enable_debug)
		return;
	va_start(args, fmt);
	vaf.fmt = fmt;
	vaf.va = &args;
	printk(KERN_DEBUG "%s: (%s, %u): %pV\n", file, func, line, &vaf);
	va_end(args);
}
EXPORT_SYMBOL(__jbd2_debug);
#endif

125
/* Checksumming functions */
126
static int jbd2_verify_csum_type(journal_t *j, journal_superblock_t *sb)
127
{
128
	if (!jbd2_journal_has_csum_v2or3_feature(j))
129 130 131 132 133
		return 1;

	return sb->s_checksum_type == JBD2_CRC32C_CHKSUM;
}

134
static __be32 jbd2_superblock_csum(journal_t *j, journal_superblock_t *sb)
135
{
136 137
	__u32 csum;
	__be32 old_csum;
138 139 140 141 142 143 144 145 146

	old_csum = sb->s_checksum;
	sb->s_checksum = 0;
	csum = jbd2_chksum(j, ~0, (char *)sb, sizeof(journal_superblock_t));
	sb->s_checksum = old_csum;

	return cpu_to_be32(csum);
}

147
static int jbd2_superblock_csum_verify(journal_t *j, journal_superblock_t *sb)
148
{
149
	if (!jbd2_journal_has_csum_v2or3(j))
150 151 152 153 154
		return 1;

	return sb->s_checksum == jbd2_superblock_csum(j, sb);
}

155
static void jbd2_superblock_csum_set(journal_t *j, journal_superblock_t *sb)
156
{
157
	if (!jbd2_journal_has_csum_v2or3(j))
158 159 160 161 162
		return;

	sb->s_checksum = jbd2_superblock_csum(j, sb);
}

163 164 165 166 167 168 169 170 171 172 173 174
/*
 * Helper function used to manage commit timeouts
 */

static void commit_timeout(unsigned long __data)
{
	struct task_struct * p = (struct task_struct *) __data;

	wake_up_process(p);
}

/*
175
 * kjournald2: The main thread function used to manage a logging device
176 177 178 179 180 181 182 183 184 185 186 187 188 189
 * journal.
 *
 * This kernel thread is responsible for two things:
 *
 * 1) COMMIT:  Every so often we need to commit the current state of the
 *    filesystem to disk.  The journal thread is responsible for writing
 *    all of the metadata buffers to disk.
 *
 * 2) CHECKPOINT: We cannot reuse a used section of the log file until all
 *    of the data in that part of the log has been rewritten elsewhere on
 *    the disk.  Flushing these old buffers to reclaim space in the log is
 *    known as checkpointing, and this thread is responsible for that job.
 */

190
static int kjournald2(void *arg)
191 192 193 194 195 196 197 198 199 200 201
{
	journal_t *journal = arg;
	transaction_t *transaction;

	/*
	 * Set up an interval timer which can be used to trigger a commit wakeup
	 * after the commit interval expires
	 */
	setup_timer(&journal->j_commit_timer, commit_timeout,
			(unsigned long)current);

202 203
	set_freezable();

204 205 206 207 208 209 210
	/* Record that the journal thread is running */
	journal->j_task = current;
	wake_up(&journal->j_wait_done_commit);

	/*
	 * And now, wait forever for commit wakeup events.
	 */
211
	write_lock(&journal->j_state_lock);
212 213

loop:
214
	if (journal->j_flags & JBD2_UNMOUNT)
215 216 217 218 219 220 221
		goto end_loop;

	jbd_debug(1, "commit_sequence=%d, commit_request=%d\n",
		journal->j_commit_sequence, journal->j_commit_request);

	if (journal->j_commit_sequence != journal->j_commit_request) {
		jbd_debug(1, "OK, requests differ\n");
222
		write_unlock(&journal->j_state_lock);
223
		del_timer_sync(&journal->j_commit_timer);
224
		jbd2_journal_commit_transaction(journal);
225
		write_lock(&journal->j_state_lock);
226 227 228 229 230 231 232 233 234 235
		goto loop;
	}

	wake_up(&journal->j_wait_done_commit);
	if (freezing(current)) {
		/*
		 * The simpler the better. Flushing journal isn't a
		 * good idea, because that depends on threads that may
		 * be already stopped.
		 */
236
		jbd_debug(1, "Now suspending kjournald2\n");
237
		write_unlock(&journal->j_state_lock);
238
		try_to_freeze();
239
		write_lock(&journal->j_state_lock);
240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255
	} else {
		/*
		 * We assume on resume that commits are already there,
		 * so we don't sleep
		 */
		DEFINE_WAIT(wait);
		int should_sleep = 1;

		prepare_to_wait(&journal->j_wait_commit, &wait,
				TASK_INTERRUPTIBLE);
		if (journal->j_commit_sequence != journal->j_commit_request)
			should_sleep = 0;
		transaction = journal->j_running_transaction;
		if (transaction && time_after_eq(jiffies,
						transaction->t_expires))
			should_sleep = 0;
256
		if (journal->j_flags & JBD2_UNMOUNT)
257 258
			should_sleep = 0;
		if (should_sleep) {
259
			write_unlock(&journal->j_state_lock);
260
			schedule();
261
			write_lock(&journal->j_state_lock);
262 263 264 265
		}
		finish_wait(&journal->j_wait_commit, &wait);
	}

266
	jbd_debug(1, "kjournald2 wakes\n");
267 268 269 270 271 272 273 274 275 276 277 278

	/*
	 * Were we woken up by a commit wakeup event?
	 */
	transaction = journal->j_running_transaction;
	if (transaction && time_after_eq(jiffies, transaction->t_expires)) {
		journal->j_commit_request = transaction->t_tid;
		jbd_debug(1, "woke because of timeout\n");
	}
	goto loop;

end_loop:
279
	write_unlock(&journal->j_state_lock);
280 281 282 283 284 285 286
	del_timer_sync(&journal->j_commit_timer);
	journal->j_task = NULL;
	wake_up(&journal->j_wait_done_commit);
	jbd_debug(1, "Journal thread exiting.\n");
	return 0;
}

287
static int jbd2_journal_start_thread(journal_t *journal)
288
{
289 290
	struct task_struct *t;

291 292
	t = kthread_run(kjournald2, journal, "jbd2/%s",
			journal->j_devname);
293 294 295
	if (IS_ERR(t))
		return PTR_ERR(t);

A
Al Viro 已提交
296
	wait_event(journal->j_wait_done_commit, journal->j_task != NULL);
297
	return 0;
298 299 300 301
}

static void journal_kill_thread(journal_t *journal)
{
302
	write_lock(&journal->j_state_lock);
303
	journal->j_flags |= JBD2_UNMOUNT;
304 305

	while (journal->j_task) {
306
		write_unlock(&journal->j_state_lock);
307
		wake_up(&journal->j_wait_commit);
A
Al Viro 已提交
308
		wait_event(journal->j_wait_done_commit, journal->j_task == NULL);
309
		write_lock(&journal->j_state_lock);
310
	}
311
	write_unlock(&journal->j_state_lock);
312 313 314
}

/*
315
 * jbd2_journal_write_metadata_buffer: write a metadata buffer to the journal.
316 317 318 319 320 321 322
 *
 * Writes a metadata buffer to a given disk block.  The actual IO is not
 * performed but a new buffer_head is constructed which labels the data
 * to be written with the correct destination disk block.
 *
 * Any magic-number escaping which needs to be done will cause a
 * copy-out here.  If the buffer happens to start with the
323
 * JBD2_MAGIC_NUMBER, then we can't write it to the log directly: the
324 325 326 327 328 329 330 331 332
 * magic number is only written to the log for descripter blocks.  In
 * this case, we copy the data and replace the first word with 0, and we
 * return a result code which indicates that this buffer needs to be
 * marked as an escaped buffer in the corresponding log descriptor
 * block.  The missing word can then be restored when the block is read
 * during recovery.
 *
 * If the source buffer has already been modified by a new transaction
 * since we took the last commit snapshot, we use the frozen copy of
333 334 335
 * that data for IO. If we end up using the existing buffer_head's data
 * for the write, then we have to make sure nobody modifies it while the
 * IO is in progress. do_get_write_access() handles this.
336
 *
337 338
 * The function returns a pointer to the buffer_head to be used for IO.
 * 
339 340 341 342 343 344 345 346 347 348
 *
 * Return value:
 *  <0: Error
 * >=0: Finished OK
 *
 * On success:
 * Bit 0 set == escape performed on the data
 * Bit 1 set == buffer copy-out performed (kfree the data after IO)
 */

349
int jbd2_journal_write_metadata_buffer(transaction_t *transaction,
350
				  struct journal_head  *jh_in,
351 352
				  struct buffer_head **bh_out,
				  sector_t blocknr)
353 354 355 356 357 358 359 360 361
{
	int need_copy_out = 0;
	int done_copy_out = 0;
	int do_escape = 0;
	char *mapped_data;
	struct buffer_head *new_bh;
	struct page *new_page;
	unsigned int new_offset;
	struct buffer_head *bh_in = jh2bh(jh_in);
362
	journal_t *journal = transaction->t_journal;
363 364 365 366 367 368 369 370 371 372 373 374

	/*
	 * The buffer really shouldn't be locked: only the current committing
	 * transaction is allowed to write it, so nobody else is allowed
	 * to do any IO.
	 *
	 * akpm: except if we're journalling data, and write() output is
	 * also part of a shared mapping, and another thread has
	 * decided to launch a writepage() against this buffer.
	 */
	J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in));

375
	new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL);
376

377 378
	/* keep subsequent assertions sane */
	atomic_set(&new_bh->b_count, 1);
379

380 381
	jbd_lock_bh_state(bh_in);
repeat:
382 383 384 385 386 387 388 389 390 391 392 393 394
	/*
	 * If a new transaction has already done a buffer copy-out, then
	 * we use that version of the data for the commit.
	 */
	if (jh_in->b_frozen_data) {
		done_copy_out = 1;
		new_page = virt_to_page(jh_in->b_frozen_data);
		new_offset = offset_in_page(jh_in->b_frozen_data);
	} else {
		new_page = jh2bh(jh_in)->b_page;
		new_offset = offset_in_page(jh2bh(jh_in)->b_data);
	}

395
	mapped_data = kmap_atomic(new_page);
J
Joel Becker 已提交
396
	/*
397 398 399 400
	 * Fire data frozen trigger if data already wasn't frozen.  Do this
	 * before checking for escaping, as the trigger may modify the magic
	 * offset.  If a copy-out happens afterwards, it will have the correct
	 * data in the buffer.
J
Joel Becker 已提交
401
	 */
402 403 404
	if (!done_copy_out)
		jbd2_buffer_frozen_trigger(jh_in, mapped_data + new_offset,
					   jh_in->b_triggers);
J
Joel Becker 已提交
405

406 407 408 409
	/*
	 * Check for escaping
	 */
	if (*((__be32 *)(mapped_data + new_offset)) ==
410
				cpu_to_be32(JBD2_MAGIC_NUMBER)) {
411 412 413
		need_copy_out = 1;
		do_escape = 1;
	}
414
	kunmap_atomic(mapped_data);
415 416 417 418 419 420 421 422

	/*
	 * Do we need to do a data copy?
	 */
	if (need_copy_out && !done_copy_out) {
		char *tmp;

		jbd_unlock_bh_state(bh_in);
M
Mingming Cao 已提交
423
		tmp = jbd2_alloc(bh_in->b_size, GFP_NOFS);
424
		if (!tmp) {
425
			brelse(new_bh);
426 427
			return -ENOMEM;
		}
428 429
		jbd_lock_bh_state(bh_in);
		if (jh_in->b_frozen_data) {
M
Mingming Cao 已提交
430
			jbd2_free(tmp, bh_in->b_size);
431 432 433 434
			goto repeat;
		}

		jh_in->b_frozen_data = tmp;
435
		mapped_data = kmap_atomic(new_page);
436
		memcpy(tmp, mapped_data + new_offset, bh_in->b_size);
437
		kunmap_atomic(mapped_data);
438 439 440 441

		new_page = virt_to_page(tmp);
		new_offset = offset_in_page(tmp);
		done_copy_out = 1;
J
Joel Becker 已提交
442 443 444 445 446 447 448

		/*
		 * This isn't strictly necessary, as we're using frozen
		 * data for the escaping, but it keeps consistency with
		 * b_frozen_data usage.
		 */
		jh_in->b_frozen_triggers = jh_in->b_triggers;
449 450 451 452 453 454 455
	}

	/*
	 * Did we need to do an escaping?  Now we've done all the
	 * copying, we can finally do so.
	 */
	if (do_escape) {
456
		mapped_data = kmap_atomic(new_page);
457
		*((unsigned int *)(mapped_data + new_offset)) = 0;
458
		kunmap_atomic(mapped_data);
459 460 461
	}

	set_bh_page(new_bh, new_page, new_offset);
462 463
	new_bh->b_size = bh_in->b_size;
	new_bh->b_bdev = journal->j_dev;
464
	new_bh->b_blocknr = blocknr;
465
	new_bh->b_private = bh_in;
466 467 468
	set_buffer_mapped(new_bh);
	set_buffer_dirty(new_bh);

469
	*bh_out = new_bh;
470 471 472 473 474 475 476

	/*
	 * The to-be-written buffer needs to get moved to the io queue,
	 * and the original buffer whose contents we are shadowing or
	 * copying is moved to the transaction's shadow queue.
	 */
	JBUFFER_TRACE(jh_in, "file as BJ_Shadow");
477 478 479
	spin_lock(&journal->j_list_lock);
	__jbd2_journal_file_buffer(jh_in, transaction, BJ_Shadow);
	spin_unlock(&journal->j_list_lock);
480
	set_buffer_shadow(bh_in);
481 482
	jbd_unlock_bh_state(bh_in);

483 484 485 486 487 488 489 490 491
	return do_escape | (done_copy_out << 1);
}

/*
 * Allocation code for the journal file.  Manage the space left in the
 * journal, so that we can begin checkpointing when appropriate.
 */

/*
492 493
 * Called with j_state_lock locked for writing.
 * Returns true if a transaction commit was started.
494
 */
495
int __jbd2_log_start_commit(journal_t *journal, tid_t target)
496
{
497 498 499 500
	/* Return if the txn has already requested to be committed */
	if (journal->j_commit_request == target)
		return 0;

501
	/*
502 503 504
	 * The only transaction we can possibly wait upon is the
	 * currently running transaction (if it exists).  Otherwise,
	 * the target tid must be an old one.
505
	 */
506 507
	if (journal->j_running_transaction &&
	    journal->j_running_transaction->t_tid == target) {
508
		/*
A
Andrea Gelmini 已提交
509
		 * We want a new commit: OK, mark the request and wakeup the
510 511 512 513
		 * commit thread.  We do _not_ do the commit ourselves.
		 */

		journal->j_commit_request = target;
E
Eryu Guan 已提交
514
		jbd_debug(1, "JBD2: requesting commit %d/%d\n",
515 516
			  journal->j_commit_request,
			  journal->j_commit_sequence);
517
		journal->j_running_transaction->t_requested = jiffies;
518 519
		wake_up(&journal->j_wait_commit);
		return 1;
520 521 522 523
	} else if (!tid_geq(journal->j_commit_request, target))
		/* This should never happen, but if it does, preserve
		   the evidence before kjournald goes into a loop and
		   increments j_commit_sequence beyond all recognition. */
E
Eryu Guan 已提交
524
		WARN_ONCE(1, "JBD2: bad log_start_commit: %u %u %u %u\n",
525 526 527 528
			  journal->j_commit_request,
			  journal->j_commit_sequence,
			  target, journal->j_running_transaction ? 
			  journal->j_running_transaction->t_tid : 0);
529 530 531
	return 0;
}

532
int jbd2_log_start_commit(journal_t *journal, tid_t tid)
533 534 535
{
	int ret;

536
	write_lock(&journal->j_state_lock);
537
	ret = __jbd2_log_start_commit(journal, tid);
538
	write_unlock(&journal->j_state_lock);
539 540 541 542
	return ret;
}

/*
543 544 545 546 547
 * Force and wait any uncommitted transactions.  We can only force the running
 * transaction if we don't have an active handle, otherwise, we will deadlock.
 * Returns: <0 in case of error,
 *           0 if nothing to commit,
 *           1 if transaction was successfully committed.
548
 */
549
static int __jbd2_journal_force_commit(journal_t *journal)
550 551 552
{
	transaction_t *transaction = NULL;
	tid_t tid;
553
	int need_to_start = 0, ret = 0;
554

555
	read_lock(&journal->j_state_lock);
556 557
	if (journal->j_running_transaction && !current->journal_info) {
		transaction = journal->j_running_transaction;
558 559
		if (!tid_geq(journal->j_commit_request, transaction->t_tid))
			need_to_start = 1;
560 561 562 563
	} else if (journal->j_committing_transaction)
		transaction = journal->j_committing_transaction;

	if (!transaction) {
564
		/* Nothing to commit */
565
		read_unlock(&journal->j_state_lock);
566
		return 0;
567 568
	}
	tid = transaction->t_tid;
569
	read_unlock(&journal->j_state_lock);
570 571
	if (need_to_start)
		jbd2_log_start_commit(journal, tid);
572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610
	ret = jbd2_log_wait_commit(journal, tid);
	if (!ret)
		ret = 1;

	return ret;
}

/**
 * Force and wait upon a commit if the calling process is not within
 * transaction.  This is used for forcing out undo-protected data which contains
 * bitmaps, when the fs is running out of space.
 *
 * @journal: journal to force
 * Returns true if progress was made.
 */
int jbd2_journal_force_commit_nested(journal_t *journal)
{
	int ret;

	ret = __jbd2_journal_force_commit(journal);
	return ret > 0;
}

/**
 * int journal_force_commit() - force any uncommitted transactions
 * @journal: journal to force
 *
 * Caller want unconditional commit. We can only force the running transaction
 * if we don't have an active handle, otherwise, we will deadlock.
 */
int jbd2_journal_force_commit(journal_t *journal)
{
	int ret;

	J_ASSERT(!current->journal_info);
	ret = __jbd2_journal_force_commit(journal);
	if (ret > 0)
		ret = 0;
	return ret;
611 612 613 614
}

/*
 * Start a commit of the current running transaction (if any).  Returns true
615 616
 * if a transaction is going to be committed (or is currently already
 * committing), and fills its tid in at *ptid
617
 */
618
int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid)
619 620 621
{
	int ret = 0;

622
	write_lock(&journal->j_state_lock);
623 624 625
	if (journal->j_running_transaction) {
		tid_t tid = journal->j_running_transaction->t_tid;

626 627 628 629
		__jbd2_log_start_commit(journal, tid);
		/* There's a running transaction and we've just made sure
		 * it's commit has been scheduled. */
		if (ptid)
630
			*ptid = tid;
631 632
		ret = 1;
	} else if (journal->j_committing_transaction) {
633
		/*
634 635
		 * If commit has been started, then we have to wait for
		 * completion of that transaction.
636
		 */
637 638
		if (ptid)
			*ptid = journal->j_committing_transaction->t_tid;
639 640
		ret = 1;
	}
641
	write_unlock(&journal->j_state_lock);
642 643 644
	return ret;
}

645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685
/*
 * Return 1 if a given transaction has not yet sent barrier request
 * connected with a transaction commit. If 0 is returned, transaction
 * may or may not have sent the barrier. Used to avoid sending barrier
 * twice in common cases.
 */
int jbd2_trans_will_send_data_barrier(journal_t *journal, tid_t tid)
{
	int ret = 0;
	transaction_t *commit_trans;

	if (!(journal->j_flags & JBD2_BARRIER))
		return 0;
	read_lock(&journal->j_state_lock);
	/* Transaction already committed? */
	if (tid_geq(journal->j_commit_sequence, tid))
		goto out;
	commit_trans = journal->j_committing_transaction;
	if (!commit_trans || commit_trans->t_tid != tid) {
		ret = 1;
		goto out;
	}
	/*
	 * Transaction is being committed and we already proceeded to
	 * submitting a flush to fs partition?
	 */
	if (journal->j_fs_dev != journal->j_dev) {
		if (!commit_trans->t_need_data_flush ||
		    commit_trans->t_state >= T_COMMIT_DFLUSH)
			goto out;
	} else {
		if (commit_trans->t_state >= T_COMMIT_JFLUSH)
			goto out;
	}
	ret = 1;
out:
	read_unlock(&journal->j_state_lock);
	return ret;
}
EXPORT_SYMBOL(jbd2_trans_will_send_data_barrier);

686 687 688 689
/*
 * Wait for a specified commit to complete.
 * The caller may not hold the journal lock.
 */
690
int jbd2_log_wait_commit(journal_t *journal, tid_t tid)
691 692 693
{
	int err = 0;

694
	jbd2_might_wait_for_commit(journal);
695
	read_lock(&journal->j_state_lock);
696
#ifdef CONFIG_JBD2_DEBUG
697
	if (!tid_geq(journal->j_commit_request, tid)) {
J
Jan Kara 已提交
698
		printk(KERN_ERR
699
		       "%s: error: j_commit_request=%d, tid=%d\n",
700
		       __func__, journal->j_commit_request, tid);
701 702 703
	}
#endif
	while (tid_gt(tid, journal->j_commit_sequence)) {
E
Eryu Guan 已提交
704
		jbd_debug(1, "JBD2: want %d, j_commit_sequence=%d\n",
705
				  tid, journal->j_commit_sequence);
706
		read_unlock(&journal->j_state_lock);
707
		wake_up(&journal->j_wait_commit);
708 709
		wait_event(journal->j_wait_done_commit,
				!tid_gt(tid, journal->j_commit_sequence));
710
		read_lock(&journal->j_state_lock);
711
	}
712
	read_unlock(&journal->j_state_lock);
713

J
Jan Kara 已提交
714
	if (unlikely(is_journal_aborted(journal)))
715 716 717 718
		err = -EIO;
	return err;
}

719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749
/*
 * When this function returns the transaction corresponding to tid
 * will be completed.  If the transaction has currently running, start
 * committing that transaction before waiting for it to complete.  If
 * the transaction id is stale, it is by definition already completed,
 * so just return SUCCESS.
 */
int jbd2_complete_transaction(journal_t *journal, tid_t tid)
{
	int	need_to_wait = 1;

	read_lock(&journal->j_state_lock);
	if (journal->j_running_transaction &&
	    journal->j_running_transaction->t_tid == tid) {
		if (journal->j_commit_request != tid) {
			/* transaction not yet started, so request it */
			read_unlock(&journal->j_state_lock);
			jbd2_log_start_commit(journal, tid);
			goto wait_commit;
		}
	} else if (!(journal->j_committing_transaction &&
		     journal->j_committing_transaction->t_tid == tid))
		need_to_wait = 0;
	read_unlock(&journal->j_state_lock);
	if (!need_to_wait)
		return 0;
wait_commit:
	return jbd2_log_wait_commit(journal, tid);
}
EXPORT_SYMBOL(jbd2_complete_transaction);

750 751 752 753
/*
 * Log buffer allocation routines:
 */

754
int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp)
755 756 757
{
	unsigned long blocknr;

758
	write_lock(&journal->j_state_lock);
759 760 761 762 763 764 765
	J_ASSERT(journal->j_free > 1);

	blocknr = journal->j_head;
	journal->j_head++;
	journal->j_free--;
	if (journal->j_head == journal->j_last)
		journal->j_head = journal->j_first;
766
	write_unlock(&journal->j_state_lock);
767
	return jbd2_journal_bmap(journal, blocknr, retp);
768 769 770 771 772 773 774 775 776
}

/*
 * Conversion of logical to physical block numbers for the journal
 *
 * On external journals the journal blocks are identity-mapped, so
 * this is a no-op.  If needed, we can use j_blk_offset - everything is
 * ready.
 */
777
int jbd2_journal_bmap(journal_t *journal, unsigned long blocknr,
778
		 unsigned long long *retp)
779 780
{
	int err = 0;
781
	unsigned long long ret;
782 783 784 785 786 787 788 789

	if (journal->j_inode) {
		ret = bmap(journal->j_inode, blocknr);
		if (ret)
			*retp = ret;
		else {
			printk(KERN_ALERT "%s: journal block not found "
					"at offset %lu on %s\n",
790
			       __func__, blocknr, journal->j_devname);
791 792 793 794 795 796 797 798 799 800 801 802 803 804
			err = -EIO;
			__journal_abort_soft(journal, err);
		}
	} else {
		*retp = blocknr; /* +journal->j_blk_offset */
	}
	return err;
}

/*
 * We play buffer_head aliasing tricks to write data/metadata blocks to
 * the journal without copying their contents, but for journal
 * descriptor blocks we do need to generate bona fide buffers.
 *
805
 * After the caller of jbd2_journal_get_descriptor_buffer() has finished modifying
806 807 808 809
 * the buffer's contents they really should run flush_dcache_page(bh->b_page).
 * But we don't bother doing that, so there will be coherency problems with
 * mmaps of blockdevs which hold live JBD-controlled filesystems.
 */
810 811
struct buffer_head *
jbd2_journal_get_descriptor_buffer(transaction_t *transaction, int type)
812
{
813
	journal_t *journal = transaction->t_journal;
814
	struct buffer_head *bh;
815
	unsigned long long blocknr;
816
	journal_header_t *header;
817 818
	int err;

819
	err = jbd2_journal_next_log_block(journal, &blocknr);
820 821 822 823 824

	if (err)
		return NULL;

	bh = __getblk(journal->j_dev, blocknr, journal->j_blocksize);
825 826
	if (!bh)
		return NULL;
827 828
	lock_buffer(bh);
	memset(bh->b_data, 0, journal->j_blocksize);
829 830 831 832
	header = (journal_header_t *)bh->b_data;
	header->h_magic = cpu_to_be32(JBD2_MAGIC_NUMBER);
	header->h_blocktype = cpu_to_be32(type);
	header->h_sequence = cpu_to_be32(transaction->t_tid);
833 834 835
	set_buffer_uptodate(bh);
	unlock_buffer(bh);
	BUFFER_TRACE(bh, "return this buffer");
836
	return bh;
837 838
}

839 840 841 842 843 844 845 846 847 848 849 850 851 852 853
void jbd2_descriptor_block_csum_set(journal_t *j, struct buffer_head *bh)
{
	struct jbd2_journal_block_tail *tail;
	__u32 csum;

	if (!jbd2_journal_has_csum_v2or3(j))
		return;

	tail = (struct jbd2_journal_block_tail *)(bh->b_data + j->j_blocksize -
			sizeof(struct jbd2_journal_block_tail));
	tail->t_checksum = 0;
	csum = jbd2_chksum(j, j->j_csum_seed, bh->b_data, j->j_blocksize);
	tail->t_checksum = cpu_to_be32(csum);
}

854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902
/*
 * Return tid of the oldest transaction in the journal and block in the journal
 * where the transaction starts.
 *
 * If the journal is now empty, return which will be the next transaction ID
 * we will write and where will that transaction start.
 *
 * The return value is 0 if journal tail cannot be pushed any further, 1 if
 * it can.
 */
int jbd2_journal_get_log_tail(journal_t *journal, tid_t *tid,
			      unsigned long *block)
{
	transaction_t *transaction;
	int ret;

	read_lock(&journal->j_state_lock);
	spin_lock(&journal->j_list_lock);
	transaction = journal->j_checkpoint_transactions;
	if (transaction) {
		*tid = transaction->t_tid;
		*block = transaction->t_log_start;
	} else if ((transaction = journal->j_committing_transaction) != NULL) {
		*tid = transaction->t_tid;
		*block = transaction->t_log_start;
	} else if ((transaction = journal->j_running_transaction) != NULL) {
		*tid = transaction->t_tid;
		*block = journal->j_head;
	} else {
		*tid = journal->j_transaction_sequence;
		*block = journal->j_head;
	}
	ret = tid_gt(*tid, journal->j_tail_sequence);
	spin_unlock(&journal->j_list_lock);
	read_unlock(&journal->j_state_lock);

	return ret;
}

/*
 * Update information in journal structure and in on disk journal superblock
 * about log tail. This function does not check whether information passed in
 * really pushes log tail further. It's responsibility of the caller to make
 * sure provided log tail information is valid (e.g. by holding
 * j_checkpoint_mutex all the time between computing log tail and calling this
 * function as is the case with jbd2_cleanup_journal_tail()).
 *
 * Requires j_checkpoint_mutex
 */
903
int __jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block)
904 905
{
	unsigned long freed;
906
	int ret;
907 908 909 910 911 912 913 914 915

	BUG_ON(!mutex_is_locked(&journal->j_checkpoint_mutex));

	/*
	 * We cannot afford for write to remain in drive's caches since as
	 * soon as we update j_tail, next transaction can start reusing journal
	 * space and if we lose sb update during power failure we'd replay
	 * old transaction with possibly newly overwritten data.
	 */
916 917 918 919
	ret = jbd2_journal_update_sb_log_tail(journal, tid, block, WRITE_FUA);
	if (ret)
		goto out;

920 921 922 923 924 925 926 927 928 929 930 931 932 933 934
	write_lock(&journal->j_state_lock);
	freed = block - journal->j_tail;
	if (block < journal->j_tail)
		freed += journal->j_last - journal->j_first;

	trace_jbd2_update_log_tail(journal, tid, block, freed);
	jbd_debug(1,
		  "Cleaning journal tail from %d to %d (offset %lu), "
		  "freeing %lu\n",
		  journal->j_tail_sequence, tid, block, freed);

	journal->j_free += freed;
	journal->j_tail_sequence = tid;
	journal->j_tail = block;
	write_unlock(&journal->j_state_lock);
935 936 937

out:
	return ret;
938 939
}

940 941 942 943 944 945 946 947 948 949 950 951 952
/*
 * This is a variaon of __jbd2_update_log_tail which checks for validity of
 * provided log tail and locks j_checkpoint_mutex. So it is safe against races
 * with other threads updating log tail.
 */
void jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block)
{
	mutex_lock(&journal->j_checkpoint_mutex);
	if (tid_gt(tid, journal->j_tail_sequence))
		__jbd2_update_log_tail(journal, tid, block);
	mutex_unlock(&journal->j_checkpoint_mutex);
}

953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975
struct jbd2_stats_proc_session {
	journal_t *journal;
	struct transaction_stats_s *stats;
	int start;
	int max;
};

static void *jbd2_seq_info_start(struct seq_file *seq, loff_t *pos)
{
	return *pos ? NULL : SEQ_START_TOKEN;
}

static void *jbd2_seq_info_next(struct seq_file *seq, void *v, loff_t *pos)
{
	return NULL;
}

static int jbd2_seq_info_show(struct seq_file *seq, void *v)
{
	struct jbd2_stats_proc_session *s = seq->private;

	if (v != SEQ_START_TOKEN)
		return 0;
976 977 978 979
	seq_printf(seq, "%lu transactions (%lu requested), "
		   "each up to %u blocks\n",
		   s->stats->ts_tid, s->stats->ts_requested,
		   s->journal->j_max_transaction_buffers);
980 981 982
	if (s->stats->ts_tid == 0)
		return 0;
	seq_printf(seq, "average: \n  %ums waiting for transaction\n",
983
	    jiffies_to_msecs(s->stats->run.rs_wait / s->stats->ts_tid));
984 985 986 987
	seq_printf(seq, "  %ums request delay\n",
	    (s->stats->ts_requested == 0) ? 0 :
	    jiffies_to_msecs(s->stats->run.rs_request_delay /
			     s->stats->ts_requested));
988
	seq_printf(seq, "  %ums running transaction\n",
989
	    jiffies_to_msecs(s->stats->run.rs_running / s->stats->ts_tid));
990
	seq_printf(seq, "  %ums transaction was being locked\n",
991
	    jiffies_to_msecs(s->stats->run.rs_locked / s->stats->ts_tid));
992
	seq_printf(seq, "  %ums flushing data (in ordered mode)\n",
993
	    jiffies_to_msecs(s->stats->run.rs_flushing / s->stats->ts_tid));
994
	seq_printf(seq, "  %ums logging transaction\n",
995
	    jiffies_to_msecs(s->stats->run.rs_logging / s->stats->ts_tid));
996 997
	seq_printf(seq, "  %lluus average transaction commit time\n",
		   div_u64(s->journal->j_average_commit_time, 1000));
998
	seq_printf(seq, "  %lu handles per transaction\n",
999
	    s->stats->run.rs_handle_count / s->stats->ts_tid);
1000
	seq_printf(seq, "  %lu blocks per transaction\n",
1001
	    s->stats->run.rs_blocks / s->stats->ts_tid);
1002
	seq_printf(seq, "  %lu logged blocks per transaction\n",
1003
	    s->stats->run.rs_blocks_logged / s->stats->ts_tid);
1004 1005 1006 1007 1008 1009 1010
	return 0;
}

static void jbd2_seq_info_stop(struct seq_file *seq, void *v)
{
}

J
James Morris 已提交
1011
static const struct seq_operations jbd2_seq_info_ops = {
1012 1013 1014 1015 1016 1017 1018 1019
	.start  = jbd2_seq_info_start,
	.next   = jbd2_seq_info_next,
	.stop   = jbd2_seq_info_stop,
	.show   = jbd2_seq_info_show,
};

static int jbd2_seq_info_open(struct inode *inode, struct file *file)
{
A
Al Viro 已提交
1020
	journal_t *journal = PDE_DATA(inode);
1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058
	struct jbd2_stats_proc_session *s;
	int rc, size;

	s = kmalloc(sizeof(*s), GFP_KERNEL);
	if (s == NULL)
		return -ENOMEM;
	size = sizeof(struct transaction_stats_s);
	s->stats = kmalloc(size, GFP_KERNEL);
	if (s->stats == NULL) {
		kfree(s);
		return -ENOMEM;
	}
	spin_lock(&journal->j_history_lock);
	memcpy(s->stats, &journal->j_stats, size);
	s->journal = journal;
	spin_unlock(&journal->j_history_lock);

	rc = seq_open(file, &jbd2_seq_info_ops);
	if (rc == 0) {
		struct seq_file *m = file->private_data;
		m->private = s;
	} else {
		kfree(s->stats);
		kfree(s);
	}
	return rc;

}

static int jbd2_seq_info_release(struct inode *inode, struct file *file)
{
	struct seq_file *seq = file->private_data;
	struct jbd2_stats_proc_session *s = seq->private;
	kfree(s->stats);
	kfree(s);
	return seq_release(inode, file);
}

1059
static const struct file_operations jbd2_seq_info_fops = {
1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070
	.owner		= THIS_MODULE,
	.open           = jbd2_seq_info_open,
	.read           = seq_read,
	.llseek         = seq_lseek,
	.release        = jbd2_seq_info_release,
};

static struct proc_dir_entry *proc_jbd2_stats;

static void jbd2_stats_proc_init(journal_t *journal)
{
1071
	journal->j_proc_entry = proc_mkdir(journal->j_devname, proc_jbd2_stats);
1072
	if (journal->j_proc_entry) {
1073 1074
		proc_create_data("info", S_IRUGO, journal->j_proc_entry,
				 &jbd2_seq_info_fops, journal);
1075 1076 1077 1078 1079 1080
	}
}

static void jbd2_stats_proc_exit(journal_t *journal)
{
	remove_proc_entry("info", journal->j_proc_entry);
1081
	remove_proc_entry(journal->j_devname, proc_jbd2_stats);
1082 1083
}

1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094
/*
 * Management for journal control blocks: functions to create and
 * destroy journal_t structures, and to initialise and read existing
 * journal blocks from disk.  */

/* First: create and setup a journal_t object in memory.  We initialise
 * very few fields yet: that has to wait until we have created the
 * journal structures from from scratch, or loaded them from disk. */

static journal_t * journal_init_common (void)
{
1095
	static struct lock_class_key jbd2_trans_commit_key;
1096 1097 1098
	journal_t *journal;
	int err;

1099
	journal = kzalloc(sizeof(*journal), GFP_KERNEL);
1100
	if (!journal)
1101
		return NULL;
1102 1103 1104 1105 1106

	init_waitqueue_head(&journal->j_wait_transaction_locked);
	init_waitqueue_head(&journal->j_wait_done_commit);
	init_waitqueue_head(&journal->j_wait_commit);
	init_waitqueue_head(&journal->j_wait_updates);
J
Jan Kara 已提交
1107
	init_waitqueue_head(&journal->j_wait_reserved);
1108 1109 1110 1111
	mutex_init(&journal->j_barrier);
	mutex_init(&journal->j_checkpoint_mutex);
	spin_lock_init(&journal->j_revoke_lock);
	spin_lock_init(&journal->j_list_lock);
1112
	rwlock_init(&journal->j_state_lock);
1113

1114
	journal->j_commit_interval = (HZ * JBD2_DEFAULT_MAX_COMMIT_AGE);
1115 1116
	journal->j_min_batch_time = 0;
	journal->j_max_batch_time = 15000; /* 15ms */
J
Jan Kara 已提交
1117
	atomic_set(&journal->j_reserved_credits, 0);
1118 1119

	/* The journal is marked for error until we succeed with recovery! */
1120
	journal->j_flags = JBD2_ABORT;
1121 1122

	/* Set up a default-sized revoke table for the new mount. */
1123
	err = jbd2_journal_init_revoke(journal, JOURNAL_REVOKE_DEFAULT_HASH);
1124 1125
	if (err) {
		kfree(journal);
1126
		return NULL;
1127
	}
1128

1129
	spin_lock_init(&journal->j_history_lock);
1130

1131 1132 1133
	lockdep_init_map(&journal->j_trans_commit_map, "jbd2_handle",
			 &jbd2_trans_commit_key, 0);

1134 1135 1136
	return journal;
}

1137
/* jbd2_journal_init_dev and jbd2_journal_init_inode:
1138 1139 1140 1141 1142 1143 1144 1145 1146
 *
 * Create a journal structure assigned some fixed set of disk blocks to
 * the journal.  We don't actually touch those disk blocks yet, but we
 * need to set up all of the mapping information to tell the journaling
 * system where the journal blocks are.
 *
 */

/**
R
Randy Dunlap 已提交
1147
 *  journal_t * jbd2_journal_init_dev() - creates and initialises a journal structure
1148 1149 1150 1151 1152
 *  @bdev: Block device on which to create the journal
 *  @fs_dev: Device which hold journalled filesystem for this journal.
 *  @start: Block nr Start of journal.
 *  @len:  Length of the journal in blocks.
 *  @blocksize: blocksize of journalling device
R
Randy Dunlap 已提交
1153 1154
 *
 *  Returns: a newly created journal_t *
1155
 *
1156
 *  jbd2_journal_init_dev creates a journal which maps a fixed contiguous
1157 1158 1159
 *  range of blocks on an arbitrary block device.
 *
 */
1160
journal_t * jbd2_journal_init_dev(struct block_device *bdev,
1161
			struct block_device *fs_dev,
1162
			unsigned long long start, int len, int blocksize)
1163 1164 1165 1166 1167 1168 1169 1170 1171 1172
{
	journal_t *journal = journal_init_common();
	struct buffer_head *bh;
	int n;

	if (!journal)
		return NULL;

	/* journal descriptor can store up to n blocks -bzzz */
	journal->j_blocksize = blocksize;
1173 1174 1175 1176 1177
	journal->j_dev = bdev;
	journal->j_fs_dev = fs_dev;
	journal->j_blk_offset = start;
	journal->j_maxlen = len;
	bdevname(journal->j_dev, journal->j_devname);
1178
	strreplace(journal->j_devname, '/', '!');
1179
	jbd2_stats_proc_init(journal);
1180 1181 1182 1183
	n = journal->j_blocksize / sizeof(journal_block_tag_t);
	journal->j_wbufsize = n;
	journal->j_wbuf = kmalloc(n * sizeof(struct buffer_head*), GFP_KERNEL);
	if (!journal->j_wbuf) {
L
Lucas De Marchi 已提交
1184
		printk(KERN_ERR "%s: Can't allocate bhs for commit thread\n",
1185
			__func__);
1186
		goto out_err;
1187 1188 1189
	}

	bh = __getblk(journal->j_dev, start, journal->j_blocksize);
1190 1191 1192 1193 1194 1195
	if (!bh) {
		printk(KERN_ERR
		       "%s: Cannot get buffer for journal superblock\n",
		       __func__);
		goto out_err;
	}
1196 1197
	journal->j_sb_buffer = bh;
	journal->j_superblock = (journal_superblock_t *)bh->b_data;
1198

1199
	return journal;
1200
out_err:
1201
	kfree(journal->j_wbuf);
1202 1203 1204
	jbd2_stats_proc_exit(journal);
	kfree(journal);
	return NULL;
1205 1206 1207
}

/**
1208
 *  journal_t * jbd2_journal_init_inode () - creates a journal which maps to a inode.
1209 1210
 *  @inode: An inode to create the journal in
 *
1211
 * jbd2_journal_init_inode creates a journal which maps an on-disk inode as
1212 1213 1214
 * the journal.  The inode must exist already, must support bmap() and
 * must have all data blocks preallocated.
 */
1215
journal_t * jbd2_journal_init_inode (struct inode *inode)
1216 1217 1218
{
	struct buffer_head *bh;
	journal_t *journal = journal_init_common();
1219
	char *p;
1220 1221
	int err;
	int n;
1222
	unsigned long long blocknr;
1223 1224 1225 1226 1227 1228

	if (!journal)
		return NULL;

	journal->j_dev = journal->j_fs_dev = inode->i_sb->s_bdev;
	journal->j_inode = inode;
1229
	bdevname(journal->j_dev, journal->j_devname);
1230
	p = strreplace(journal->j_devname, '/', '!');
1231
	sprintf(p, "-%lu", journal->j_inode->i_ino);
1232 1233 1234 1235 1236 1237 1238 1239
	jbd_debug(1,
		  "journal %p: inode %s/%ld, size %Ld, bits %d, blksize %ld\n",
		  journal, inode->i_sb->s_id, inode->i_ino,
		  (long long) inode->i_size,
		  inode->i_sb->s_blocksize_bits, inode->i_sb->s_blocksize);

	journal->j_maxlen = inode->i_size >> inode->i_sb->s_blocksize_bits;
	journal->j_blocksize = inode->i_sb->s_blocksize;
1240
	jbd2_stats_proc_init(journal);
1241 1242 1243 1244 1245 1246

	/* journal descriptor can store up to n blocks -bzzz */
	n = journal->j_blocksize / sizeof(journal_block_tag_t);
	journal->j_wbufsize = n;
	journal->j_wbuf = kmalloc(n * sizeof(struct buffer_head*), GFP_KERNEL);
	if (!journal->j_wbuf) {
L
Lucas De Marchi 已提交
1247
		printk(KERN_ERR "%s: Can't allocate bhs for commit thread\n",
1248
			__func__);
1249
		goto out_err;
1250 1251
	}

1252
	err = jbd2_journal_bmap(journal, 0, &blocknr);
1253 1254
	/* If that failed, give up */
	if (err) {
1255
		printk(KERN_ERR "%s: Cannot locate journal superblock\n",
1256
		       __func__);
1257
		goto out_err;
1258 1259
	}

1260
	bh = getblk_unmovable(journal->j_dev, blocknr, journal->j_blocksize);
1261 1262 1263 1264 1265 1266
	if (!bh) {
		printk(KERN_ERR
		       "%s: Cannot get buffer for journal superblock\n",
		       __func__);
		goto out_err;
	}
1267 1268 1269 1270
	journal->j_sb_buffer = bh;
	journal->j_superblock = (journal_superblock_t *)bh->b_data;

	return journal;
1271
out_err:
1272
	kfree(journal->j_wbuf);
1273 1274 1275
	jbd2_stats_proc_exit(journal);
	kfree(journal);
	return NULL;
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299
}

/*
 * If the journal init or create aborts, we need to mark the journal
 * superblock as being NULL to prevent the journal destroy from writing
 * back a bogus superblock.
 */
static void journal_fail_superblock (journal_t *journal)
{
	struct buffer_head *bh = journal->j_sb_buffer;
	brelse(bh);
	journal->j_sb_buffer = NULL;
}

/*
 * Given a journal_t structure, initialise the various fields for
 * startup of a new journaling session.  We use this both when creating
 * a journal, and after recovering an old journal to reset it for
 * subsequent use.
 */

static int journal_reset(journal_t *journal)
{
	journal_superblock_t *sb = journal->j_superblock;
1300
	unsigned long long first, last;
1301 1302 1303

	first = be32_to_cpu(sb->s_first);
	last = be32_to_cpu(sb->s_maxlen);
1304
	if (first + JBD2_MIN_JOURNAL_BLOCKS > last + 1) {
E
Eryu Guan 已提交
1305
		printk(KERN_ERR "JBD2: Journal too short (blocks %llu-%llu).\n",
1306 1307 1308 1309
		       first, last);
		journal_fail_superblock(journal);
		return -EINVAL;
	}
1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325

	journal->j_first = first;
	journal->j_last = last;

	journal->j_head = first;
	journal->j_tail = first;
	journal->j_free = last - first;

	journal->j_tail_sequence = journal->j_transaction_sequence;
	journal->j_commit_sequence = journal->j_transaction_sequence - 1;
	journal->j_commit_request = journal->j_commit_sequence;

	journal->j_max_transaction_buffers = journal->j_maxlen / 4;

	/*
	 * As a special case, if the on-disk copy is already marked as needing
1326 1327
	 * no recovery (s_start == 0), then we can safely defer the superblock
	 * update until the next commit by setting JBD2_FLUSHED.  This avoids
1328 1329
	 * attempting a write to a potential-readonly device.
	 */
1330
	if (sb->s_start == 0) {
E
Eryu Guan 已提交
1331
		jbd_debug(1, "JBD2: Skipping superblock update on recovered sb "
1332 1333 1334
			"(start %ld, seq %d, errno %d)\n",
			journal->j_tail, journal->j_tail_sequence,
			journal->j_errno);
1335 1336
		journal->j_flags |= JBD2_FLUSHED;
	} else {
1337 1338
		/* Lock here to make assertions happy... */
		mutex_lock(&journal->j_checkpoint_mutex);
1339 1340 1341 1342 1343 1344 1345 1346 1347 1348
		/*
		 * Update log tail information. We use WRITE_FUA since new
		 * transaction will start reusing journal space and so we
		 * must make sure information about current log tail is on
		 * disk before that.
		 */
		jbd2_journal_update_sb_log_tail(journal,
						journal->j_tail_sequence,
						journal->j_tail,
						WRITE_FUA);
1349
		mutex_unlock(&journal->j_checkpoint_mutex);
1350
	}
1351 1352
	return jbd2_journal_start_thread(journal);
}
1353

1354
static int jbd2_write_superblock(journal_t *journal, int write_flags)
1355 1356
{
	struct buffer_head *bh = journal->j_sb_buffer;
1357
	journal_superblock_t *sb = journal->j_superblock;
1358
	int ret;
1359

1360
	trace_jbd2_write_superblock(journal, write_flags);
1361
	if (!(journal->j_flags & JBD2_BARRIER))
1362
		write_flags &= ~(REQ_FUA | REQ_PREFLUSH);
1363
	lock_buffer(bh);
1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378
	if (buffer_write_io_error(bh)) {
		/*
		 * Oh, dear.  A previous attempt to write the journal
		 * superblock failed.  This could happen because the
		 * USB device was yanked out.  Or it could happen to
		 * be a transient write error and maybe the block will
		 * be remapped.  Nothing we can do but to retry the
		 * write and hope for the best.
		 */
		printk(KERN_ERR "JBD2: previous I/O error detected "
		       "for journal superblock update for %s.\n",
		       journal->j_devname);
		clear_buffer_write_io_error(bh);
		set_buffer_uptodate(bh);
	}
1379
	jbd2_superblock_csum_set(journal, sb);
1380 1381
	get_bh(bh);
	bh->b_end_io = end_buffer_write_sync;
1382
	ret = submit_bh(REQ_OP_WRITE, write_flags, bh);
1383
	wait_on_buffer(bh);
1384 1385 1386
	if (buffer_write_io_error(bh)) {
		clear_buffer_write_io_error(bh);
		set_buffer_uptodate(bh);
1387 1388 1389 1390 1391 1392
		ret = -EIO;
	}
	if (ret) {
		printk(KERN_ERR "JBD2: Error %d detected when updating "
		       "journal superblock for %s.\n", ret,
		       journal->j_devname);
1393
		jbd2_journal_abort(journal, ret);
1394
	}
1395 1396

	return ret;
1397 1398 1399 1400 1401
}

/**
 * jbd2_journal_update_sb_log_tail() - Update log tail in journal sb on disk.
 * @journal: The journal to update.
1402 1403 1404
 * @tail_tid: TID of the new transaction at the tail of the log
 * @tail_block: The first block of the transaction at the tail of the log
 * @write_op: With which operation should we write the journal sb
1405 1406 1407 1408
 *
 * Update a journal's superblock information about log tail and write it to
 * disk, waiting for the IO to complete.
 */
1409
int jbd2_journal_update_sb_log_tail(journal_t *journal, tid_t tail_tid,
1410
				     unsigned long tail_block, int write_op)
1411 1412
{
	journal_superblock_t *sb = journal->j_superblock;
1413
	int ret;
1414

1415
	BUG_ON(!mutex_is_locked(&journal->j_checkpoint_mutex));
1416 1417
	jbd_debug(1, "JBD2: updating superblock (start %lu, seq %u)\n",
		  tail_block, tail_tid);
1418

1419 1420
	sb->s_sequence = cpu_to_be32(tail_tid);
	sb->s_start    = cpu_to_be32(tail_block);
1421

1422 1423 1424
	ret = jbd2_write_superblock(journal, write_op);
	if (ret)
		goto out;
1425

1426 1427 1428 1429 1430
	/* Log is no longer empty */
	write_lock(&journal->j_state_lock);
	WARN_ON(!sb->s_sequence);
	journal->j_flags &= ~JBD2_FLUSHED;
	write_unlock(&journal->j_state_lock);
1431 1432 1433

out:
	return ret;
1434
}
1435

1436 1437 1438
/**
 * jbd2_mark_journal_empty() - Mark on disk journal as empty.
 * @journal: The journal to update.
1439
 * @write_op: With which operation should we write the journal sb
1440 1441 1442 1443
 *
 * Update a journal's dynamic superblock fields to show that journal is empty.
 * Write updated superblock to disk waiting for IO to complete.
 */
1444
static void jbd2_mark_journal_empty(journal_t *journal, int write_op)
1445 1446
{
	journal_superblock_t *sb = journal->j_superblock;
1447

1448
	BUG_ON(!mutex_is_locked(&journal->j_checkpoint_mutex));
1449
	read_lock(&journal->j_state_lock);
1450 1451 1452 1453 1454
	/* Is it already empty? */
	if (sb->s_start == 0) {
		read_unlock(&journal->j_state_lock);
		return;
	}
1455 1456
	jbd_debug(1, "JBD2: Marking journal as empty (seq %d)\n",
		  journal->j_tail_sequence);
1457 1458

	sb->s_sequence = cpu_to_be32(journal->j_tail_sequence);
1459
	sb->s_start    = cpu_to_be32(0);
1460
	read_unlock(&journal->j_state_lock);
1461

1462
	jbd2_write_superblock(journal, write_op);
1463

1464
	/* Log is no longer empty */
1465
	write_lock(&journal->j_state_lock);
1466
	journal->j_flags |= JBD2_FLUSHED;
1467
	write_unlock(&journal->j_state_lock);
1468 1469
}

1470 1471 1472 1473 1474 1475 1476 1477

/**
 * jbd2_journal_update_sb_errno() - Update error in the journal.
 * @journal: The journal to update.
 *
 * Update a journal's errno.  Write updated superblock to disk waiting for IO
 * to complete.
 */
1478
void jbd2_journal_update_sb_errno(journal_t *journal)
1479 1480 1481 1482 1483 1484 1485 1486 1487
{
	journal_superblock_t *sb = journal->j_superblock;

	read_lock(&journal->j_state_lock);
	jbd_debug(1, "JBD2: updating superblock error (errno %d)\n",
		  journal->j_errno);
	sb->s_errno    = cpu_to_be32(journal->j_errno);
	read_unlock(&journal->j_state_lock);

1488
	jbd2_write_superblock(journal, WRITE_FUA);
1489
}
1490
EXPORT_SYMBOL(jbd2_journal_update_sb_errno);
1491

1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505
/*
 * Read the superblock for a given journal, performing initial
 * validation of the format.
 */
static int journal_get_superblock(journal_t *journal)
{
	struct buffer_head *bh;
	journal_superblock_t *sb;
	int err = -EIO;

	bh = journal->j_sb_buffer;

	J_ASSERT(bh != NULL);
	if (!buffer_uptodate(bh)) {
1506
		ll_rw_block(REQ_OP_READ, 0, 1, &bh);
1507 1508
		wait_on_buffer(bh);
		if (!buffer_uptodate(bh)) {
E
Eryu Guan 已提交
1509 1510
			printk(KERN_ERR
				"JBD2: IO error reading journal superblock\n");
1511 1512 1513 1514
			goto out;
		}
	}

1515 1516 1517
	if (buffer_verified(bh))
		return 0;

1518 1519 1520 1521
	sb = journal->j_superblock;

	err = -EINVAL;

1522
	if (sb->s_header.h_magic != cpu_to_be32(JBD2_MAGIC_NUMBER) ||
1523
	    sb->s_blocksize != cpu_to_be32(journal->j_blocksize)) {
E
Eryu Guan 已提交
1524
		printk(KERN_WARNING "JBD2: no valid journal superblock found\n");
1525 1526 1527 1528
		goto out;
	}

	switch(be32_to_cpu(sb->s_header.h_blocktype)) {
1529
	case JBD2_SUPERBLOCK_V1:
1530 1531
		journal->j_format_version = 1;
		break;
1532
	case JBD2_SUPERBLOCK_V2:
1533 1534 1535
		journal->j_format_version = 2;
		break;
	default:
E
Eryu Guan 已提交
1536
		printk(KERN_WARNING "JBD2: unrecognised superblock format ID\n");
1537 1538 1539 1540 1541 1542
		goto out;
	}

	if (be32_to_cpu(sb->s_maxlen) < journal->j_maxlen)
		journal->j_maxlen = be32_to_cpu(sb->s_maxlen);
	else if (be32_to_cpu(sb->s_maxlen) > journal->j_maxlen) {
E
Eryu Guan 已提交
1543
		printk(KERN_WARNING "JBD2: journal file too short\n");
1544 1545 1546
		goto out;
	}

1547 1548 1549 1550 1551 1552 1553 1554
	if (be32_to_cpu(sb->s_first) == 0 ||
	    be32_to_cpu(sb->s_first) >= journal->j_maxlen) {
		printk(KERN_WARNING
			"JBD2: Invalid start block of journal: %u\n",
			be32_to_cpu(sb->s_first));
		goto out;
	}

1555 1556
	if (jbd2_has_feature_csum2(journal) &&
	    jbd2_has_feature_csum3(journal)) {
1557 1558 1559 1560 1561 1562
		/* Can't have checksum v2 and v3 at the same time! */
		printk(KERN_ERR "JBD2: Can't enable checksumming v2 and v3 "
		       "at the same time!\n");
		goto out;
	}

1563
	if (jbd2_journal_has_csum_v2or3_feature(journal) &&
1564
	    jbd2_has_feature_checksum(journal)) {
1565 1566 1567 1568 1569 1570
		/* Can't have checksum v1 and v2 on at the same time! */
		printk(KERN_ERR "JBD2: Can't enable checksumming v1 and v2/3 "
		       "at the same time!\n");
		goto out;
	}

1571
	if (!jbd2_verify_csum_type(journal, sb)) {
1572
		printk(KERN_ERR "JBD2: Unknown checksum type\n");
1573 1574 1575
		goto out;
	}

1576
	/* Load the checksum driver */
1577
	if (jbd2_journal_has_csum_v2or3_feature(journal)) {
1578 1579
		journal->j_chksum_driver = crypto_alloc_shash("crc32c", 0, 0);
		if (IS_ERR(journal->j_chksum_driver)) {
1580
			printk(KERN_ERR "JBD2: Cannot load crc32c driver.\n");
1581 1582 1583 1584 1585 1586
			err = PTR_ERR(journal->j_chksum_driver);
			journal->j_chksum_driver = NULL;
			goto out;
		}
	}

1587 1588
	/* Check superblock checksum */
	if (!jbd2_superblock_csum_verify(journal, sb)) {
1589
		printk(KERN_ERR "JBD2: journal checksum error\n");
1590
		err = -EFSBADCRC;
1591 1592 1593 1594
		goto out;
	}

	/* Precompute checksum seed for all metadata */
1595
	if (jbd2_journal_has_csum_v2or3(journal))
1596 1597 1598
		journal->j_csum_seed = jbd2_chksum(journal, ~0, sb->s_uuid,
						   sizeof(sb->s_uuid));

1599 1600
	set_buffer_verified(bh);

1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634
	return 0;

out:
	journal_fail_superblock(journal);
	return err;
}

/*
 * Load the on-disk journal superblock and read the key fields into the
 * journal_t.
 */

static int load_superblock(journal_t *journal)
{
	int err;
	journal_superblock_t *sb;

	err = journal_get_superblock(journal);
	if (err)
		return err;

	sb = journal->j_superblock;

	journal->j_tail_sequence = be32_to_cpu(sb->s_sequence);
	journal->j_tail = be32_to_cpu(sb->s_start);
	journal->j_first = be32_to_cpu(sb->s_first);
	journal->j_last = be32_to_cpu(sb->s_maxlen);
	journal->j_errno = be32_to_cpu(sb->s_errno);

	return 0;
}


/**
1635
 * int jbd2_journal_load() - Read journal from disk.
1636 1637 1638 1639 1640 1641
 * @journal: Journal to act on.
 *
 * Given a journal_t structure which tells us which disk blocks contain
 * a journal, read the journal from disk to initialise the in-memory
 * structures.
 */
1642
int jbd2_journal_load(journal_t *journal)
1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656
{
	int err;
	journal_superblock_t *sb;

	err = load_superblock(journal);
	if (err)
		return err;

	sb = journal->j_superblock;
	/* If this is a V2 superblock, then we have to check the
	 * features flags on it. */

	if (journal->j_format_version >= 2) {
		if ((sb->s_feature_ro_compat &
1657
		     ~cpu_to_be32(JBD2_KNOWN_ROCOMPAT_FEATURES)) ||
1658
		    (sb->s_feature_incompat &
1659
		     ~cpu_to_be32(JBD2_KNOWN_INCOMPAT_FEATURES))) {
E
Eryu Guan 已提交
1660 1661
			printk(KERN_WARNING
				"JBD2: Unrecognised features on journal\n");
1662 1663 1664 1665
			return -EINVAL;
		}
	}

1666 1667 1668 1669 1670 1671 1672
	/*
	 * Create a slab for this blocksize
	 */
	err = jbd2_journal_create_slab(be32_to_cpu(sb->s_blocksize));
	if (err)
		return err;

1673 1674
	/* Let the recovery code check whether it needs to recover any
	 * data from the journal. */
1675
	if (jbd2_journal_recover(journal))
1676 1677
		goto recovery_error;

1678 1679 1680 1681
	if (journal->j_failed_commit) {
		printk(KERN_ERR "JBD2: journal transaction %u on %s "
		       "is corrupt.\n", journal->j_failed_commit,
		       journal->j_devname);
1682
		return -EFSCORRUPTED;
1683 1684
	}

1685 1686 1687 1688 1689 1690
	/* OK, we've finished with the dynamic journal bits:
	 * reinitialise the dynamic contents of the superblock in memory
	 * and reset them on disk. */
	if (journal_reset(journal))
		goto recovery_error;

1691 1692
	journal->j_flags &= ~JBD2_ABORT;
	journal->j_flags |= JBD2_LOADED;
1693 1694 1695
	return 0;

recovery_error:
E
Eryu Guan 已提交
1696
	printk(KERN_WARNING "JBD2: recovery failed\n");
1697 1698 1699 1700
	return -EIO;
}

/**
1701
 * void jbd2_journal_destroy() - Release a journal_t structure.
1702 1703 1704 1705
 * @journal: Journal to act on.
 *
 * Release a journal_t structure once it is no longer in use by the
 * journaled object.
1706
 * Return <0 if we couldn't clean up the journal.
1707
 */
1708
int jbd2_journal_destroy(journal_t *journal)
1709
{
1710 1711
	int err = 0;

1712 1713 1714 1715 1716
	/* Wait for the commit thread to wake up and die. */
	journal_kill_thread(journal);

	/* Force a final log commit */
	if (journal->j_running_transaction)
1717
		jbd2_journal_commit_transaction(journal);
1718 1719 1720 1721 1722 1723 1724

	/* Force any old transactions to disk */

	/* Totally anal locking here... */
	spin_lock(&journal->j_list_lock);
	while (journal->j_checkpoint_transactions != NULL) {
		spin_unlock(&journal->j_list_lock);
1725
		mutex_lock(&journal->j_checkpoint_mutex);
1726
		err = jbd2_log_do_checkpoint(journal);
1727
		mutex_unlock(&journal->j_checkpoint_mutex);
1728 1729 1730 1731 1732 1733 1734 1735 1736
		/*
		 * If checkpointing failed, just free the buffers to avoid
		 * looping forever
		 */
		if (err) {
			jbd2_journal_destroy_checkpoint(journal);
			spin_lock(&journal->j_list_lock);
			break;
		}
1737 1738 1739 1740 1741 1742 1743 1744 1745
		spin_lock(&journal->j_list_lock);
	}

	J_ASSERT(journal->j_running_transaction == NULL);
	J_ASSERT(journal->j_committing_transaction == NULL);
	J_ASSERT(journal->j_checkpoint_transactions == NULL);
	spin_unlock(&journal->j_list_lock);

	if (journal->j_sb_buffer) {
1746
		if (!is_journal_aborted(journal)) {
1747
			mutex_lock(&journal->j_checkpoint_mutex);
1748 1749 1750 1751 1752 1753 1754

			write_lock(&journal->j_state_lock);
			journal->j_tail_sequence =
				++journal->j_transaction_sequence;
			write_unlock(&journal->j_state_lock);

			jbd2_mark_journal_empty(journal, WRITE_FLUSH_FUA);
1755 1756
			mutex_unlock(&journal->j_checkpoint_mutex);
		} else
1757
			err = -EIO;
1758 1759 1760
		brelse(journal->j_sb_buffer);
	}

1761 1762
	if (journal->j_proc_entry)
		jbd2_stats_proc_exit(journal);
1763
	iput(journal->j_inode);
1764
	if (journal->j_revoke)
1765
		jbd2_journal_destroy_revoke(journal);
1766 1767
	if (journal->j_chksum_driver)
		crypto_free_shash(journal->j_chksum_driver);
1768 1769
	kfree(journal->j_wbuf);
	kfree(journal);
1770 1771

	return err;
1772 1773 1774 1775
}


/**
1776
 *int jbd2_journal_check_used_features () - Check if features specified are used.
1777 1778 1779 1780 1781 1782 1783 1784 1785
 * @journal: Journal to check.
 * @compat: bitmask of compatible features
 * @ro: bitmask of features that force read-only mount
 * @incompat: bitmask of incompatible features
 *
 * Check whether the journal uses all of a given set of
 * features.  Return true (non-zero) if it does.
 **/

1786
int jbd2_journal_check_used_features (journal_t *journal, unsigned long compat,
1787 1788 1789 1790 1791 1792
				 unsigned long ro, unsigned long incompat)
{
	journal_superblock_t *sb;

	if (!compat && !ro && !incompat)
		return 1;
1793 1794 1795 1796
	/* Load journal superblock if it is not loaded yet. */
	if (journal->j_format_version == 0 &&
	    journal_get_superblock(journal) != 0)
		return 0;
1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810
	if (journal->j_format_version == 1)
		return 0;

	sb = journal->j_superblock;

	if (((be32_to_cpu(sb->s_feature_compat) & compat) == compat) &&
	    ((be32_to_cpu(sb->s_feature_ro_compat) & ro) == ro) &&
	    ((be32_to_cpu(sb->s_feature_incompat) & incompat) == incompat))
		return 1;

	return 0;
}

/**
1811
 * int jbd2_journal_check_available_features() - Check feature set in journalling layer
1812 1813 1814 1815 1816 1817 1818 1819 1820
 * @journal: Journal to check.
 * @compat: bitmask of compatible features
 * @ro: bitmask of features that force read-only mount
 * @incompat: bitmask of incompatible features
 *
 * Check whether the journaling code supports the use of
 * all of a given set of features on this journal.  Return true
 * (non-zero) if it can. */

1821
int jbd2_journal_check_available_features (journal_t *journal, unsigned long compat,
1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833
				      unsigned long ro, unsigned long incompat)
{
	if (!compat && !ro && !incompat)
		return 1;

	/* We can support any known requested features iff the
	 * superblock is in version 2.  Otherwise we fail to support any
	 * extended sb features. */

	if (journal->j_format_version != 2)
		return 0;

1834 1835 1836
	if ((compat   & JBD2_KNOWN_COMPAT_FEATURES) == compat &&
	    (ro       & JBD2_KNOWN_ROCOMPAT_FEATURES) == ro &&
	    (incompat & JBD2_KNOWN_INCOMPAT_FEATURES) == incompat)
1837 1838 1839 1840 1841 1842
		return 1;

	return 0;
}

/**
1843
 * int jbd2_journal_set_features () - Mark a given journal feature in the superblock
1844 1845 1846 1847 1848 1849 1850 1851 1852 1853
 * @journal: Journal to act on.
 * @compat: bitmask of compatible features
 * @ro: bitmask of features that force read-only mount
 * @incompat: bitmask of incompatible features
 *
 * Mark a given journal feature as present on the
 * superblock.  Returns true if the requested features could be set.
 *
 */

1854
int jbd2_journal_set_features (journal_t *journal, unsigned long compat,
1855 1856
			  unsigned long ro, unsigned long incompat)
{
1857 1858 1859 1860
#define INCOMPAT_FEATURE_ON(f) \
		((incompat & (f)) && !(sb->s_feature_incompat & cpu_to_be32(f)))
#define COMPAT_FEATURE_ON(f) \
		((compat & (f)) && !(sb->s_feature_compat & cpu_to_be32(f)))
1861 1862
	journal_superblock_t *sb;

1863
	if (jbd2_journal_check_used_features(journal, compat, ro, incompat))
1864 1865
		return 1;

1866
	if (!jbd2_journal_check_available_features(journal, compat, ro, incompat))
1867 1868
		return 0;

1869 1870 1871 1872 1873 1874 1875 1876
	/* If enabling v2 checksums, turn on v3 instead */
	if (incompat & JBD2_FEATURE_INCOMPAT_CSUM_V2) {
		incompat &= ~JBD2_FEATURE_INCOMPAT_CSUM_V2;
		incompat |= JBD2_FEATURE_INCOMPAT_CSUM_V3;
	}

	/* Asking for checksumming v3 and v1?  Only give them v3. */
	if (incompat & JBD2_FEATURE_INCOMPAT_CSUM_V3 &&
1877 1878 1879
	    compat & JBD2_FEATURE_COMPAT_CHECKSUM)
		compat &= ~JBD2_FEATURE_COMPAT_CHECKSUM;

1880 1881 1882 1883 1884
	jbd_debug(1, "Setting new features 0x%lx/0x%lx/0x%lx\n",
		  compat, ro, incompat);

	sb = journal->j_superblock;

1885 1886
	/* If enabling v3 checksums, update superblock */
	if (INCOMPAT_FEATURE_ON(JBD2_FEATURE_INCOMPAT_CSUM_V3)) {
1887 1888 1889
		sb->s_checksum_type = JBD2_CRC32C_CHKSUM;
		sb->s_feature_compat &=
			~cpu_to_be32(JBD2_FEATURE_COMPAT_CHECKSUM);
1890 1891 1892 1893 1894 1895

		/* Load the checksum driver */
		if (journal->j_chksum_driver == NULL) {
			journal->j_chksum_driver = crypto_alloc_shash("crc32c",
								      0, 0);
			if (IS_ERR(journal->j_chksum_driver)) {
1896
				printk(KERN_ERR "JBD2: Cannot load crc32c "
1897 1898 1899 1900
				       "driver.\n");
				journal->j_chksum_driver = NULL;
				return 0;
			}
1901

1902
			/* Precompute checksum seed for all metadata */
1903 1904 1905
			journal->j_csum_seed = jbd2_chksum(journal, ~0,
							   sb->s_uuid,
							   sizeof(sb->s_uuid));
1906
		}
1907 1908 1909 1910 1911
	}

	/* If enabling v1 checksums, downgrade superblock */
	if (COMPAT_FEATURE_ON(JBD2_FEATURE_COMPAT_CHECKSUM))
		sb->s_feature_incompat &=
1912 1913
			~cpu_to_be32(JBD2_FEATURE_INCOMPAT_CSUM_V2 |
				     JBD2_FEATURE_INCOMPAT_CSUM_V3);
1914

1915 1916 1917 1918 1919
	sb->s_feature_compat    |= cpu_to_be32(compat);
	sb->s_feature_ro_compat |= cpu_to_be32(ro);
	sb->s_feature_incompat  |= cpu_to_be32(incompat);

	return 1;
1920 1921
#undef COMPAT_FEATURE_ON
#undef INCOMPAT_FEATURE_ON
1922 1923
}

1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949
/*
 * jbd2_journal_clear_features () - Clear a given journal feature in the
 * 				    superblock
 * @journal: Journal to act on.
 * @compat: bitmask of compatible features
 * @ro: bitmask of features that force read-only mount
 * @incompat: bitmask of incompatible features
 *
 * Clear a given journal feature as present on the
 * superblock.
 */
void jbd2_journal_clear_features(journal_t *journal, unsigned long compat,
				unsigned long ro, unsigned long incompat)
{
	journal_superblock_t *sb;

	jbd_debug(1, "Clear features 0x%lx/0x%lx/0x%lx\n",
		  compat, ro, incompat);

	sb = journal->j_superblock;

	sb->s_feature_compat    &= ~cpu_to_be32(compat);
	sb->s_feature_ro_compat &= ~cpu_to_be32(ro);
	sb->s_feature_incompat  &= ~cpu_to_be32(incompat);
}
EXPORT_SYMBOL(jbd2_journal_clear_features);
1950 1951

/**
1952
 * int jbd2_journal_flush () - Flush journal
1953 1954 1955 1956 1957 1958 1959
 * @journal: Journal to act on.
 *
 * Flush all data for a given journal to disk and empty the journal.
 * Filesystems can use this when remounting readonly to ensure that
 * recovery does not need to happen on remount.
 */

1960
int jbd2_journal_flush(journal_t *journal)
1961 1962 1963 1964
{
	int err = 0;
	transaction_t *transaction = NULL;

1965
	write_lock(&journal->j_state_lock);
1966 1967 1968 1969

	/* Force everything buffered to the log... */
	if (journal->j_running_transaction) {
		transaction = journal->j_running_transaction;
1970
		__jbd2_log_start_commit(journal, transaction->t_tid);
1971 1972 1973 1974 1975 1976 1977
	} else if (journal->j_committing_transaction)
		transaction = journal->j_committing_transaction;

	/* Wait for the log commit to complete... */
	if (transaction) {
		tid_t tid = transaction->t_tid;

1978
		write_unlock(&journal->j_state_lock);
1979
		jbd2_log_wait_commit(journal, tid);
1980
	} else {
1981
		write_unlock(&journal->j_state_lock);
1982 1983 1984 1985 1986 1987
	}

	/* ...and flush everything in the log out to disk. */
	spin_lock(&journal->j_list_lock);
	while (!err && journal->j_checkpoint_transactions != NULL) {
		spin_unlock(&journal->j_list_lock);
1988
		mutex_lock(&journal->j_checkpoint_mutex);
1989
		err = jbd2_log_do_checkpoint(journal);
1990
		mutex_unlock(&journal->j_checkpoint_mutex);
1991 1992 1993
		spin_lock(&journal->j_list_lock);
	}
	spin_unlock(&journal->j_list_lock);
1994 1995 1996 1997

	if (is_journal_aborted(journal))
		return -EIO;

1998
	mutex_lock(&journal->j_checkpoint_mutex);
1999 2000 2001 2002 2003 2004 2005 2006
	if (!err) {
		err = jbd2_cleanup_journal_tail(journal);
		if (err < 0) {
			mutex_unlock(&journal->j_checkpoint_mutex);
			goto out;
		}
		err = 0;
	}
2007 2008 2009 2010 2011 2012

	/* Finally, mark the journal as really needing no recovery.
	 * This sets s_start==0 in the underlying superblock, which is
	 * the magic code for a fully-recovered superblock.  Any future
	 * commits of data to the journal will restore the current
	 * s_start value. */
2013
	jbd2_mark_journal_empty(journal, WRITE_FUA);
2014
	mutex_unlock(&journal->j_checkpoint_mutex);
2015
	write_lock(&journal->j_state_lock);
2016 2017 2018 2019 2020
	J_ASSERT(!journal->j_running_transaction);
	J_ASSERT(!journal->j_committing_transaction);
	J_ASSERT(!journal->j_checkpoint_transactions);
	J_ASSERT(journal->j_head == journal->j_tail);
	J_ASSERT(journal->j_tail_sequence == journal->j_transaction_sequence);
2021
	write_unlock(&journal->j_state_lock);
2022 2023
out:
	return err;
2024 2025 2026
}

/**
2027
 * int jbd2_journal_wipe() - Wipe journal contents
2028 2029 2030 2031 2032
 * @journal: Journal to act on.
 * @write: flag (see below)
 *
 * Wipe out all of the contents of a journal, safely.  This will produce
 * a warning if the journal contains any valid recovery information.
2033
 * Must be called between journal_init_*() and jbd2_journal_load().
2034 2035 2036 2037 2038
 *
 * If 'write' is non-zero, then we wipe out the journal on disk; otherwise
 * we merely suppress recovery.
 */

2039
int jbd2_journal_wipe(journal_t *journal, int write)
2040 2041 2042
{
	int err = 0;

2043
	J_ASSERT (!(journal->j_flags & JBD2_LOADED));
2044 2045 2046 2047 2048 2049 2050 2051

	err = load_superblock(journal);
	if (err)
		return err;

	if (!journal->j_tail)
		goto no_recovery;

E
Eryu Guan 已提交
2052
	printk(KERN_WARNING "JBD2: %s recovery information on journal\n",
2053 2054
		write ? "Clearing" : "Ignoring");

2055
	err = jbd2_journal_skip_recovery(journal);
2056 2057 2058
	if (write) {
		/* Lock to make assertions happy... */
		mutex_lock(&journal->j_checkpoint_mutex);
2059
		jbd2_mark_journal_empty(journal, WRITE_FUA);
2060 2061
		mutex_unlock(&journal->j_checkpoint_mutex);
	}
2062 2063 2064 2065 2066 2067 2068 2069 2070

 no_recovery:
	return err;
}

/*
 * Journal abort has very specific semantics, which we describe
 * for journal abort.
 *
2071
 * Two internal functions, which provide abort to the jbd layer
2072 2073 2074 2075 2076 2077 2078 2079
 * itself are here.
 */

/*
 * Quick version for internal journal use (doesn't lock the journal).
 * Aborts hard --- we mark the abort as occurred, but do _nothing_ else,
 * and don't attempt to make any other journal updates.
 */
2080
void __jbd2_journal_abort_hard(journal_t *journal)
2081 2082 2083
{
	transaction_t *transaction;

2084
	if (journal->j_flags & JBD2_ABORT)
2085 2086 2087
		return;

	printk(KERN_ERR "Aborting journal on device %s.\n",
2088
	       journal->j_devname);
2089

2090
	write_lock(&journal->j_state_lock);
2091
	journal->j_flags |= JBD2_ABORT;
2092 2093
	transaction = journal->j_running_transaction;
	if (transaction)
2094
		__jbd2_log_start_commit(journal, transaction->t_tid);
2095
	write_unlock(&journal->j_state_lock);
2096 2097 2098 2099 2100 2101
}

/* Soft abort: record the abort error status in the journal superblock,
 * but don't do any other IO. */
static void __journal_abort_soft (journal_t *journal, int errno)
{
2102
	if (journal->j_flags & JBD2_ABORT)
2103 2104 2105 2106 2107
		return;

	if (!journal->j_errno)
		journal->j_errno = errno;

2108
	__jbd2_journal_abort_hard(journal);
2109

2110
	if (errno) {
2111
		jbd2_journal_update_sb_errno(journal);
2112 2113 2114 2115
		write_lock(&journal->j_state_lock);
		journal->j_flags |= JBD2_REC_ERR;
		write_unlock(&journal->j_state_lock);
	}
2116 2117 2118
}

/**
2119
 * void jbd2_journal_abort () - Shutdown the journal immediately.
2120 2121 2122 2123 2124 2125 2126 2127
 * @journal: the journal to shutdown.
 * @errno:   an error number to record in the journal indicating
 *           the reason for the shutdown.
 *
 * Perform a complete, immediate shutdown of the ENTIRE
 * journal (not of a single transaction).  This operation cannot be
 * undone without closing and reopening the journal.
 *
2128
 * The jbd2_journal_abort function is intended to support higher level error
2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143
 * recovery mechanisms such as the ext2/ext3 remount-readonly error
 * mode.
 *
 * Journal abort has very specific semantics.  Any existing dirty,
 * unjournaled buffers in the main filesystem will still be written to
 * disk by bdflush, but the journaling mechanism will be suspended
 * immediately and no further transaction commits will be honoured.
 *
 * Any dirty, journaled buffers will be written back to disk without
 * hitting the journal.  Atomicity cannot be guaranteed on an aborted
 * filesystem, but we _do_ attempt to leave as much data as possible
 * behind for fsck to use for cleanup.
 *
 * Any attempt to get a new transaction handle on a journal which is in
 * ABORT state will just result in an -EROFS error return.  A
2144
 * jbd2_journal_stop on an existing handle will return -EIO if we have
2145 2146 2147
 * entered abort state during the update.
 *
 * Recursive transactions are not disturbed by journal abort until the
2148
 * final jbd2_journal_stop, which will receive the -EIO error.
2149
 *
2150
 * Finally, the jbd2_journal_abort call allows the caller to supply an errno
2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163
 * which will be recorded (if possible) in the journal superblock.  This
 * allows a client to record failure conditions in the middle of a
 * transaction without having to complete the transaction to record the
 * failure to disk.  ext3_error, for example, now uses this
 * functionality.
 *
 * Errors which originate from within the journaling layer will NOT
 * supply an errno; a null errno implies that absolutely no further
 * writes are done to the journal (unless there are any already in
 * progress).
 *
 */

2164
void jbd2_journal_abort(journal_t *journal, int errno)
2165 2166 2167 2168 2169
{
	__journal_abort_soft(journal, errno);
}

/**
2170
 * int jbd2_journal_errno () - returns the journal's error state.
2171 2172
 * @journal: journal to examine.
 *
2173
 * This is the errno number set with jbd2_journal_abort(), the last
2174 2175 2176 2177 2178 2179
 * time the journal was mounted - if the journal was stopped
 * without calling abort this will be 0.
 *
 * If the journal has been aborted on this mount time -EROFS will
 * be returned.
 */
2180
int jbd2_journal_errno(journal_t *journal)
2181 2182 2183
{
	int err;

2184
	read_lock(&journal->j_state_lock);
2185
	if (journal->j_flags & JBD2_ABORT)
2186 2187 2188
		err = -EROFS;
	else
		err = journal->j_errno;
2189
	read_unlock(&journal->j_state_lock);
2190 2191 2192 2193
	return err;
}

/**
2194
 * int jbd2_journal_clear_err () - clears the journal's error state
2195 2196
 * @journal: journal to act on.
 *
2197
 * An error must be cleared or acked to take a FS out of readonly
2198 2199
 * mode.
 */
2200
int jbd2_journal_clear_err(journal_t *journal)
2201 2202 2203
{
	int err = 0;

2204
	write_lock(&journal->j_state_lock);
2205
	if (journal->j_flags & JBD2_ABORT)
2206 2207 2208
		err = -EROFS;
	else
		journal->j_errno = 0;
2209
	write_unlock(&journal->j_state_lock);
2210 2211 2212 2213
	return err;
}

/**
2214
 * void jbd2_journal_ack_err() - Ack journal err.
2215 2216
 * @journal: journal to act on.
 *
2217
 * An error must be cleared or acked to take a FS out of readonly
2218 2219
 * mode.
 */
2220
void jbd2_journal_ack_err(journal_t *journal)
2221
{
2222
	write_lock(&journal->j_state_lock);
2223
	if (journal->j_errno)
2224
		journal->j_flags |= JBD2_ACK_ERR;
2225
	write_unlock(&journal->j_state_lock);
2226 2227
}

2228
int jbd2_journal_blocks_per_page(struct inode *inode)
2229
{
2230
	return 1 << (PAGE_SHIFT - inode->i_sb->s_blocksize_bits);
2231 2232
}

Z
Zach Brown 已提交
2233 2234 2235 2236 2237
/*
 * helper functions to deal with 32 or 64bit block numbers.
 */
size_t journal_tag_bytes(journal_t *journal)
{
2238 2239
	size_t sz;

2240
	if (jbd2_has_feature_csum3(journal))
2241 2242 2243
		return sizeof(journal_block_tag3_t);

	sz = sizeof(journal_block_tag_t);
2244

2245
	if (jbd2_has_feature_csum2(journal))
2246
		sz += sizeof(__u16);
2247

2248
	if (jbd2_has_feature_64bit(journal))
2249
		return sz;
Z
Zach Brown 已提交
2250
	else
2251
		return sz - sizeof(__u32);
Z
Zach Brown 已提交
2252 2253
}

2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290
/*
 * JBD memory management
 *
 * These functions are used to allocate block-sized chunks of memory
 * used for making copies of buffer_head data.  Very often it will be
 * page-sized chunks of data, but sometimes it will be in
 * sub-page-size chunks.  (For example, 16k pages on Power systems
 * with a 4k block file system.)  For blocks smaller than a page, we
 * use a SLAB allocator.  There are slab caches for each block size,
 * which are allocated at mount time, if necessary, and we only free
 * (all of) the slab caches when/if the jbd2 module is unloaded.  For
 * this reason we don't need to a mutex to protect access to
 * jbd2_slab[] allocating or releasing memory; only in
 * jbd2_journal_create_slab().
 */
#define JBD2_MAX_SLABS 8
static struct kmem_cache *jbd2_slab[JBD2_MAX_SLABS];

static const char *jbd2_slab_names[JBD2_MAX_SLABS] = {
	"jbd2_1k", "jbd2_2k", "jbd2_4k", "jbd2_8k",
	"jbd2_16k", "jbd2_32k", "jbd2_64k", "jbd2_128k"
};


static void jbd2_journal_destroy_slabs(void)
{
	int i;

	for (i = 0; i < JBD2_MAX_SLABS; i++) {
		if (jbd2_slab[i])
			kmem_cache_destroy(jbd2_slab[i]);
		jbd2_slab[i] = NULL;
	}
}

static int jbd2_journal_create_slab(size_t size)
{
2291
	static DEFINE_MUTEX(jbd2_slab_create_mutex);
2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302
	int i = order_base_2(size) - 10;
	size_t slab_size;

	if (size == PAGE_SIZE)
		return 0;

	if (i >= JBD2_MAX_SLABS)
		return -EINVAL;

	if (unlikely(i < 0))
		i = 0;
2303
	mutex_lock(&jbd2_slab_create_mutex);
2304
	if (jbd2_slab[i]) {
2305
		mutex_unlock(&jbd2_slab_create_mutex);
2306 2307 2308 2309 2310 2311
		return 0;	/* Already created */
	}

	slab_size = 1 << (i+10);
	jbd2_slab[i] = kmem_cache_create(jbd2_slab_names[i], slab_size,
					 slab_size, 0, NULL);
2312
	mutex_unlock(&jbd2_slab_create_mutex);
2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326
	if (!jbd2_slab[i]) {
		printk(KERN_EMERG "JBD2: no memory for jbd2_slab cache\n");
		return -ENOMEM;
	}
	return 0;
}

static struct kmem_cache *get_slab(size_t size)
{
	int i = order_base_2(size) - 10;

	BUG_ON(i >= JBD2_MAX_SLABS);
	if (unlikely(i < 0))
		i = 0;
2327
	BUG_ON(jbd2_slab[i] == NULL);
2328 2329 2330 2331 2332 2333 2334 2335 2336
	return jbd2_slab[i];
}

void *jbd2_alloc(size_t size, gfp_t flags)
{
	void *ptr;

	BUG_ON(size & (size-1)); /* Must be a power of 2 */

2337
	if (size < PAGE_SIZE)
2338
		ptr = kmem_cache_alloc(get_slab(size), flags);
2339 2340
	else
		ptr = (void *)__get_free_pages(flags, get_order(size));
2341 2342 2343 2344 2345 2346 2347 2348 2349 2350

	/* Check alignment; SLUB has gotten this wrong in the past,
	 * and this can lead to user data corruption! */
	BUG_ON(((unsigned long) ptr) & (size-1));

	return ptr;
}

void jbd2_free(void *ptr, size_t size)
{
2351 2352 2353 2354
	if (size < PAGE_SIZE)
		kmem_cache_free(get_slab(size), ptr);
	else
		free_pages((unsigned long)ptr, get_order(size));
2355 2356
};

2357 2358 2359
/*
 * Journal_head storage management
 */
2360
static struct kmem_cache *jbd2_journal_head_cache;
2361
#ifdef CONFIG_JBD2_DEBUG
2362 2363 2364
static atomic_t nr_journal_heads = ATOMIC_INIT(0);
#endif

2365
static int jbd2_journal_init_journal_head_cache(void)
2366 2367 2368
{
	int retval;

A
Al Viro 已提交
2369
	J_ASSERT(jbd2_journal_head_cache == NULL);
J
Johann Lombardi 已提交
2370
	jbd2_journal_head_cache = kmem_cache_create("jbd2_journal_head",
2371 2372
				sizeof(struct journal_head),
				0,		/* offset */
2373
				SLAB_TEMPORARY | SLAB_DESTROY_BY_RCU,
2374
				NULL);		/* ctor */
2375
	retval = 0;
A
Al Viro 已提交
2376
	if (!jbd2_journal_head_cache) {
2377
		retval = -ENOMEM;
E
Eryu Guan 已提交
2378
		printk(KERN_EMERG "JBD2: no memory for journal_head cache\n");
2379 2380 2381 2382
	}
	return retval;
}

2383
static void jbd2_journal_destroy_journal_head_cache(void)
2384
{
2385 2386 2387 2388
	if (jbd2_journal_head_cache) {
		kmem_cache_destroy(jbd2_journal_head_cache);
		jbd2_journal_head_cache = NULL;
	}
2389 2390 2391 2392 2393 2394 2395 2396 2397
}

/*
 * journal_head splicing and dicing
 */
static struct journal_head *journal_alloc_journal_head(void)
{
	struct journal_head *ret;

2398
#ifdef CONFIG_JBD2_DEBUG
2399 2400
	atomic_inc(&nr_journal_heads);
#endif
2401
	ret = kmem_cache_zalloc(jbd2_journal_head_cache, GFP_NOFS);
A
Al Viro 已提交
2402
	if (!ret) {
2403
		jbd_debug(1, "out of memory for journal_head\n");
2404
		pr_notice_ratelimited("ENOMEM in %s, retrying.\n", __func__);
2405 2406
		ret = kmem_cache_zalloc(jbd2_journal_head_cache,
				GFP_NOFS | __GFP_NOFAIL);
2407 2408 2409 2410 2411 2412
	}
	return ret;
}

static void journal_free_journal_head(struct journal_head *jh)
{
2413
#ifdef CONFIG_JBD2_DEBUG
2414
	atomic_dec(&nr_journal_heads);
2415
	memset(jh, JBD2_POISON_FREE, sizeof(*jh));
2416
#endif
2417
	kmem_cache_free(jbd2_journal_head_cache, jh);
2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433
}

/*
 * A journal_head is attached to a buffer_head whenever JBD has an
 * interest in the buffer.
 *
 * Whenever a buffer has an attached journal_head, its ->b_state:BH_JBD bit
 * is set.  This bit is tested in core kernel code where we need to take
 * JBD-specific actions.  Testing the zeroness of ->b_private is not reliable
 * there.
 *
 * When a buffer has its BH_JBD bit set, its ->b_count is elevated by one.
 *
 * When a buffer has its BH_JBD bit set it is immune from being released by
 * core kernel code, mainly via ->b_count.
 *
2434 2435 2436
 * A journal_head is detached from its buffer_head when the journal_head's
 * b_jcount reaches zero. Running transaction (b_transaction) and checkpoint
 * transaction (b_cp_transaction) hold their references to b_jcount.
2437 2438 2439
 *
 * Various places in the kernel want to attach a journal_head to a buffer_head
 * _before_ attaching the journal_head to a transaction.  To protect the
2440
 * journal_head in this situation, jbd2_journal_add_journal_head elevates the
2441
 * journal_head's b_jcount refcount by one.  The caller must call
2442
 * jbd2_journal_put_journal_head() to undo this.
2443 2444 2445 2446
 *
 * So the typical usage would be:
 *
 *	(Attach a journal_head if needed.  Increments b_jcount)
2447
 *	struct journal_head *jh = jbd2_journal_add_journal_head(bh);
2448
 *	...
2449 2450
 *      (Get another reference for transaction)
 *	jbd2_journal_grab_journal_head(bh);
2451
 *	jh->b_transaction = xxx;
2452
 *	(Put original reference)
2453
 *	jbd2_journal_put_journal_head(jh);
2454 2455 2456 2457 2458 2459 2460
 */

/*
 * Give a buffer_head a journal_head.
 *
 * May sleep.
 */
2461
struct journal_head *jbd2_journal_add_journal_head(struct buffer_head *bh)
2462 2463 2464 2465 2466
{
	struct journal_head *jh;
	struct journal_head *new_jh = NULL;

repeat:
2467
	if (!buffer_jbd(bh))
2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501
		new_jh = journal_alloc_journal_head();

	jbd_lock_bh_journal_head(bh);
	if (buffer_jbd(bh)) {
		jh = bh2jh(bh);
	} else {
		J_ASSERT_BH(bh,
			(atomic_read(&bh->b_count) > 0) ||
			(bh->b_page && bh->b_page->mapping));

		if (!new_jh) {
			jbd_unlock_bh_journal_head(bh);
			goto repeat;
		}

		jh = new_jh;
		new_jh = NULL;		/* We consumed it */
		set_buffer_jbd(bh);
		bh->b_private = jh;
		jh->b_bh = bh;
		get_bh(bh);
		BUFFER_TRACE(bh, "added journal_head");
	}
	jh->b_jcount++;
	jbd_unlock_bh_journal_head(bh);
	if (new_jh)
		journal_free_journal_head(new_jh);
	return bh->b_private;
}

/*
 * Grab a ref against this buffer_head's journal_head.  If it ended up not
 * having a journal_head, return NULL
 */
2502
struct journal_head *jbd2_journal_grab_journal_head(struct buffer_head *bh)
2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519
{
	struct journal_head *jh = NULL;

	jbd_lock_bh_journal_head(bh);
	if (buffer_jbd(bh)) {
		jh = bh2jh(bh);
		jh->b_jcount++;
	}
	jbd_unlock_bh_journal_head(bh);
	return jh;
}

static void __journal_remove_journal_head(struct buffer_head *bh)
{
	struct journal_head *jh = bh2jh(bh);

	J_ASSERT_JH(jh, jh->b_jcount >= 0);
2520 2521 2522 2523 2524 2525 2526 2527 2528 2529
	J_ASSERT_JH(jh, jh->b_transaction == NULL);
	J_ASSERT_JH(jh, jh->b_next_transaction == NULL);
	J_ASSERT_JH(jh, jh->b_cp_transaction == NULL);
	J_ASSERT_JH(jh, jh->b_jlist == BJ_None);
	J_ASSERT_BH(bh, buffer_jbd(bh));
	J_ASSERT_BH(bh, jh2bh(jh) == bh);
	BUFFER_TRACE(bh, "remove journal_head");
	if (jh->b_frozen_data) {
		printk(KERN_WARNING "%s: freeing b_frozen_data\n", __func__);
		jbd2_free(jh->b_frozen_data, bh->b_size);
2530
	}
2531 2532 2533 2534 2535 2536 2537 2538
	if (jh->b_committed_data) {
		printk(KERN_WARNING "%s: freeing b_committed_data\n", __func__);
		jbd2_free(jh->b_committed_data, bh->b_size);
	}
	bh->b_private = NULL;
	jh->b_bh = NULL;	/* debug, really */
	clear_buffer_jbd(bh);
	journal_free_journal_head(jh);
2539 2540 2541
}

/*
2542
 * Drop a reference on the passed journal_head.  If it fell to zero then
2543 2544
 * release the journal_head from the buffer_head.
 */
2545
void jbd2_journal_put_journal_head(struct journal_head *jh)
2546 2547 2548 2549 2550 2551
{
	struct buffer_head *bh = jh2bh(jh);

	jbd_lock_bh_journal_head(bh);
	J_ASSERT_JH(jh, jh->b_jcount > 0);
	--jh->b_jcount;
2552
	if (!jh->b_jcount) {
2553
		__journal_remove_journal_head(bh);
2554
		jbd_unlock_bh_journal_head(bh);
2555
		__brelse(bh);
2556 2557
	} else
		jbd_unlock_bh_journal_head(bh);
2558 2559
}

2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584
/*
 * Initialize jbd inode head
 */
void jbd2_journal_init_jbd_inode(struct jbd2_inode *jinode, struct inode *inode)
{
	jinode->i_transaction = NULL;
	jinode->i_next_transaction = NULL;
	jinode->i_vfs_inode = inode;
	jinode->i_flags = 0;
	INIT_LIST_HEAD(&jinode->i_list);
}

/*
 * Function to be called before we start removing inode from memory (i.e.,
 * clear_inode() is a fine place to be called from). It removes inode from
 * transaction's lists.
 */
void jbd2_journal_release_jbd_inode(journal_t *journal,
				    struct jbd2_inode *jinode)
{
	if (!journal)
		return;
restart:
	spin_lock(&journal->j_list_lock);
	/* Is commit writing out inode - we have to wait */
2585
	if (jinode->i_flags & JI_COMMIT_RUNNING) {
2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602
		wait_queue_head_t *wq;
		DEFINE_WAIT_BIT(wait, &jinode->i_flags, __JI_COMMIT_RUNNING);
		wq = bit_waitqueue(&jinode->i_flags, __JI_COMMIT_RUNNING);
		prepare_to_wait(wq, &wait.wait, TASK_UNINTERRUPTIBLE);
		spin_unlock(&journal->j_list_lock);
		schedule();
		finish_wait(wq, &wait.wait);
		goto restart;
	}

	if (jinode->i_transaction) {
		list_del(&jinode->i_list);
		jinode->i_transaction = NULL;
	}
	spin_unlock(&journal->j_list_lock);
}

2603

2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625
#ifdef CONFIG_PROC_FS

#define JBD2_STATS_PROC_NAME "fs/jbd2"

static void __init jbd2_create_jbd_stats_proc_entry(void)
{
	proc_jbd2_stats = proc_mkdir(JBD2_STATS_PROC_NAME, NULL);
}

static void __exit jbd2_remove_jbd_stats_proc_entry(void)
{
	if (proc_jbd2_stats)
		remove_proc_entry(JBD2_STATS_PROC_NAME, NULL);
}

#else

#define jbd2_create_jbd_stats_proc_entry() do {} while (0)
#define jbd2_remove_jbd_stats_proc_entry() do {} while (0)

#endif

2626
struct kmem_cache *jbd2_handle_cache, *jbd2_inode_cache;
2627

2628
static int __init jbd2_journal_init_handle_cache(void)
2629
{
2630
	jbd2_handle_cache = KMEM_CACHE(jbd2_journal_handle, SLAB_TEMPORARY);
2631
	if (jbd2_handle_cache == NULL) {
2632 2633 2634 2635 2636 2637 2638
		printk(KERN_EMERG "JBD2: failed to create handle cache\n");
		return -ENOMEM;
	}
	jbd2_inode_cache = KMEM_CACHE(jbd2_inode, 0);
	if (jbd2_inode_cache == NULL) {
		printk(KERN_EMERG "JBD2: failed to create inode cache\n");
		kmem_cache_destroy(jbd2_handle_cache);
2639 2640 2641 2642 2643
		return -ENOMEM;
	}
	return 0;
}

2644
static void jbd2_journal_destroy_handle_cache(void)
2645
{
2646 2647
	if (jbd2_handle_cache)
		kmem_cache_destroy(jbd2_handle_cache);
2648 2649 2650
	if (jbd2_inode_cache)
		kmem_cache_destroy(jbd2_inode_cache);

2651 2652 2653 2654 2655 2656 2657 2658 2659 2660
}

/*
 * Module startup and shutdown
 */

static int __init journal_init_caches(void)
{
	int ret;

2661
	ret = jbd2_journal_init_revoke_caches();
2662
	if (ret == 0)
2663
		ret = jbd2_journal_init_journal_head_cache();
2664
	if (ret == 0)
2665
		ret = jbd2_journal_init_handle_cache();
2666
	if (ret == 0)
2667
		ret = jbd2_journal_init_transaction_cache();
2668 2669 2670
	return ret;
}

2671
static void jbd2_journal_destroy_caches(void)
2672
{
2673
	jbd2_journal_destroy_revoke_caches();
2674
	jbd2_journal_destroy_journal_head_cache();
2675
	jbd2_journal_destroy_handle_cache();
2676
	jbd2_journal_destroy_transaction_cache();
2677
	jbd2_journal_destroy_slabs();
2678 2679 2680 2681 2682 2683 2684 2685 2686
}

static int __init journal_init(void)
{
	int ret;

	BUILD_BUG_ON(sizeof(struct journal_superblock_s) != 1024);

	ret = journal_init_caches();
2687 2688 2689
	if (ret == 0) {
		jbd2_create_jbd_stats_proc_entry();
	} else {
2690
		jbd2_journal_destroy_caches();
2691
	}
2692 2693 2694 2695 2696
	return ret;
}

static void __exit journal_exit(void)
{
2697
#ifdef CONFIG_JBD2_DEBUG
2698 2699
	int n = atomic_read(&nr_journal_heads);
	if (n)
J
Jan Kara 已提交
2700
		printk(KERN_ERR "JBD2: leaked %d journal_heads!\n", n);
2701
#endif
2702
	jbd2_remove_jbd_stats_proc_entry();
2703
	jbd2_journal_destroy_caches();
2704 2705 2706 2707 2708 2709
}

MODULE_LICENSE("GPL");
module_init(journal_init);
module_exit(journal_exit);