Release 4.10 fs/jbd2/commit.c

/*
 * linux/fs/jbd2/commit.c
 *
 * 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.
 *
 * Journal commit routines for the generic filesystem journaling code;
 * part of the ext2fs journaling system.
 */

#include <linux/time.h>
#include <linux/fs.h>
#include <linux/jbd2.h>
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/mm.h>
#include <linux/pagemap.h>
#include <linux/jiffies.h>
#include <linux/crc32.h>
#include <linux/writeback.h>
#include <linux/backing-dev.h>
#include <linux/bio.h>
#include <linux/blkdev.h>
#include <linux/bitops.h>
#include <trace/events/jbd2.h>

/*
 * IO end handler for temporary buffer_heads handling writes to the journal.
 */


static void journal_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
{
	struct buffer_head *orig_bh = bh->b_private;

	BUFFER_TRACE(bh, "");
	if (uptodate)
		set_buffer_uptodate(bh);
	else
		clear_buffer_uptodate(bh);
	if (orig_bh) {
		clear_bit_unlock(BH_Shadow, &orig_bh->b_state);
		smp_mb__after_atomic();
		wake_up_bit(&orig_bh->b_state, BH_Shadow);
	}
	unlock_buffer(bh);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
dave kleikamp
dave kleikamp
41
51.90%
1
33.33%
jan kara
jan kara
37
46.84%
1
33.33%
peter zijlstra
peter zijlstra
1
1.27%
1
33.33%
Total
79
100.00%
3
100.00%

/*
 * When an ext4 file is truncated, it is possible that some pages are not
 * successfully freed, because they are attached to a committing transaction.
 * After the transaction commits, these pages are left on the LRU, with no
 * ->mapping, and with attached buffers.  These pages are trivially reclaimable
 * by the VM, but their apparent absence upsets the VM accounting, and it makes
 * the numbers in /proc/meminfo look odd.
 *
 * So here, we have a buffer which has just come off the forget list.  Look to
 * see if we can strip all buffers from the backing page.
 *
 * Called under lock_journal(), and possibly under journal_datalist_lock.  The
 * caller provided us with a ref against the buffer, and we drop that here.
 */


static void release_buffer_page(struct buffer_head *bh)
{
	struct page *page;

	if (buffer_dirty(bh))
		goto nope;
	if (atomic_read(&bh->b_count) != 1)
		goto nope;
	page = bh->b_page;
	if (!page)
		goto nope;
	if (page->mapping)
		goto nope;

	/* OK, it's a truncated page */
	if (!trylock_page(page))
		goto nope;

	get_page(page);
	__brelse(bh);
	try_to_free_buffers(page);
	unlock_page(page);
	put_page(page);
	return;

nope:
	__brelse(bh);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
dave kleikamp
dave kleikamp
105
96.33%
1
33.33%
nick piggin
nick piggin
2
1.83%
1
33.33%
kirill a. shutemov
kirill a. shutemov
2
1.83%
1
33.33%
Total
109
100.00%
3
100.00%



static void jbd2_commit_block_csum_set(journal_t *j, struct buffer_head *bh)
{
	struct commit_header *h;
	__u32 csum;

	if (!jbd2_journal_has_csum_v2or3(j))
		return;

	h = (struct commit_header *)(bh->b_data);
	h->h_chksum_type = 0;
	h->h_chksum_size = 0;
	h->h_chksum[0] = 0;
	csum = jbd2_chksum(j, j->j_csum_seed, bh->b_data, j->j_blocksize);
	h->h_chksum[0] = cpu_to_be32(csum);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
darrick j. wong
darrick j. wong
93
95.88%
2
66.67%
jan kara
jan kara
4
4.12%
1
33.33%
Total
97
100.00%
3
100.00%

/*
 * Done it all: now submit the commit record.  We should have
 * cleaned up our previous buffers by now, so if we are in abort
 * mode we can now just skip the rest of the journal write
 * entirely.
 *
 * Returns 1 if the journal needs to be aborted or 0 on success
 */


static int journal_submit_commit_record(journal_t *journal,
					transaction_t *commit_transaction,
					struct buffer_head **cbh,
					__u32 crc32_sum)
{
	struct commit_header *tmp;
	struct buffer_head *bh;
	int ret;
	struct timespec64 now = current_kernel_time64();

	*cbh = NULL;

	if (is_journal_aborted(journal))
		return 0;

	bh = jbd2_journal_get_descriptor_buffer(commit_transaction,
						JBD2_COMMIT_BLOCK);
	if (!bh)
		return 1;

	tmp = (struct commit_header *)bh->b_data;
	tmp->h_commit_sec = cpu_to_be64(now.tv_sec);
	tmp->h_commit_nsec = cpu_to_be32(now.tv_nsec);

	if (jbd2_has_feature_checksum(journal)) {
		tmp->h_chksum_type 	= JBD2_CRC32_CHKSUM;
		tmp->h_chksum_size 	= JBD2_CRC32_CHKSUM_SIZE;
		tmp->h_chksum[0] 	= cpu_to_be32(crc32_sum);
	}
	jbd2_commit_block_csum_set(journal, bh);

	BUFFER_TRACE(bh, "submit commit block");
	lock_buffer(bh);
	clear_buffer_dirty(bh);
	set_buffer_uptodate(bh);
	bh->b_end_io = journal_end_buffer_io_sync;

	if (journal->j_flags & JBD2_BARRIER &&
	    !jbd2_has_feature_async_commit(journal))
		ret = submit_bh(REQ_OP_WRITE,
			REQ_SYNC | REQ_PREFLUSH | REQ_FUA, bh);
	else
		ret = submit_bh(REQ_OP_WRITE, REQ_SYNC, bh);

	*cbh = bh;
	return ret;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
dave kleikamp
dave kleikamp
74
32.89%
1
6.67%
girish shilamkar
girish shilamkar
73
32.44%
1
6.67%
theodore tso
theodore tso
26
11.56%
2
13.33%
jan kara
jan kara
16
7.11%
3
20.00%
christoph hellwig
christoph hellwig
15
6.67%
2
13.33%
darrick j. wong
darrick j. wong
8
3.56%
2
13.33%
zhang huan
zhang huan
5
2.22%
1
6.67%
michael christie
michael christie
4
1.78%
1
6.67%
arnd bergmann
arnd bergmann
2
0.89%
1
6.67%
mingming cao
mingming cao
2
0.89%
1
6.67%
Total
225
100.00%
15
100.00%

/*
 * This function along with journal_submit_commit_record
 * allows to write the commit record asynchronously.
 */


static int journal_wait_on_commit_record(journal_t *journal,
					 struct buffer_head *bh)
{
	int ret = 0;

	clear_buffer_dirty(bh);
	wait_on_buffer(bh);

	if (unlikely(!buffer_uptodate(bh)))
		ret = -EIO;
	put_bh(bh);            /* One for getblk() */

	return ret;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
jan kara
jan kara
33
60.00%
1
33.33%
dave kleikamp
dave kleikamp
18
32.73%
1
33.33%
theodore tso
theodore tso
4
7.27%
1
33.33%
Total
55
100.00%
3
100.00%

/*
 * write the filemap data using writepage() address_space_operations.
 * We don't do block allocation here even for delalloc. We don't
 * use writepages() because with dealyed allocation we may be doing
 * block allocation in writepages().
 */


static int journal_submit_inode_data_buffers(struct address_space *mapping)
{
	int ret;
	struct writeback_control wbc = {
		.sync_mode =  WB_SYNC_ALL,
		.nr_to_write = mapping->nrpages * 2,
		.range_start = 0,
		.range_end = i_size_read(mapping->host),
        };

	ret = generic_writepages(mapping, &wbc);
	return ret;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
aneesh kumar
aneesh kumar
58
93.55%
1
50.00%
jan kara
jan kara
4
6.45%
1
50.00%
Total
62
100.00%
2
100.00%

/*
 * Submit all the data buffers of inode associated with the transaction to
 * disk.
 *
 * We are in a committing transaction. Therefore no new inode can be added to
 * our inode list. We use JI_COMMIT_RUNNING flag to protect inode we currently
 * operate on from being released while we write out pages.
 */


static int journal_submit_data_buffers(journal_t *journal,
		transaction_t *commit_transaction)
{
	struct jbd2_inode *jinode;
	int err, ret = 0;
	struct address_space *mapping;

	spin_lock(&journal->j_list_lock);
	list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) {
		if (!(jinode->i_flags & JI_WRITE_DATA))
			continue;
		mapping = jinode->i_vfs_inode->i_mapping;
		jinode->i_flags |= JI_COMMIT_RUNNING;
		spin_unlock(&journal->j_list_lock);
		/*
                 * submit the inode data buffers. We use writepage
                 * instead of writepages. Because writepages can do
                 * block allocation  with delalloc. We need to write
                 * only allocated blocks here.
                 */
		trace_jbd2_submit_inode_data(jinode->i_vfs_inode);
		err = journal_submit_inode_data_buffers(mapping);
		if (!ret)
			ret = err;
		spin_lock(&journal->j_list_lock);
		J_ASSERT(jinode->i_transaction == commit_transaction);
		jinode->i_flags &= ~JI_COMMIT_RUNNING;
		smp_mb();
		wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING);
	}
	spin_unlock(&journal->j_list_lock);
	return ret;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
jan kara
jan kara
140
90.32%
3
50.00%
theodore tso
theodore tso
7
4.52%
1
16.67%
aneesh kumar
aneesh kumar
6
3.87%
1
16.67%
brian king
brian king
2
1.29%
1
16.67%
Total
155
100.00%
6
100.00%

/*
 * Wait for data submitted for writeout, refile inodes to proper
 * transaction if needed.
 *
 */


static int journal_finish_inode_data_buffers(journal_t *journal,
		transaction_t *commit_transaction)
{
	struct jbd2_inode *jinode, *next_i;
	int err, ret = 0;

	/* For locking, see the comment in journal_submit_data_buffers() */
	spin_lock(&journal->j_list_lock);
	list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) {
		if (!(jinode->i_flags & JI_WAIT_DATA))
			continue;
		jinode->i_flags |= JI_COMMIT_RUNNING;
		spin_unlock(&journal->j_list_lock);
		err = filemap_fdatawait(jinode->i_vfs_inode->i_mapping);
		if (err) {
			/*
                         * Because AS_EIO is cleared by
                         * filemap_fdatawait_range(), set it again so
                         * that user process can get -EIO from fsync().
                         */
			mapping_set_error(jinode->i_vfs_inode->i_mapping, -EIO);

			if (!ret)
				ret = err;
		}
		spin_lock(&journal->j_list_lock);
		jinode->i_flags &= ~JI_COMMIT_RUNNING;
		smp_mb();
		wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING);
	}

	/* Now refile inode to proper lists */
	list_for_each_entry_safe(jinode, next_i,
				 &commit_transaction->t_inode_list, i_list) {
		list_del(&jinode->i_list);
		if (jinode->i_next_transaction) {
			jinode->i_transaction = jinode->i_next_transaction;
			jinode->i_next_transaction = NULL;
			list_add(&jinode->i_list,
				&jinode->i_transaction->t_inode_list);
		} else {
			jinode->i_transaction = NULL;
		}
	}
	spin_unlock(&journal->j_list_lock);

	return ret;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
jan kara
jan kara
197
89.95%
3
37.50%
hidehiro kawai
hidehiro kawai
14
6.39%
1
12.50%
michal hocko
michal hocko
4
1.83%
1
12.50%
brian king
brian king
2
0.91%
1
12.50%
aneesh kumar
aneesh kumar
1
0.46%
1
12.50%
christoph hellwig
christoph hellwig
1
0.46%
1
12.50%
Total
219
100.00%
8
100.00%



static __u32 jbd2_checksum_data(__u32 crc32_sum, struct buffer_head *bh)
{
	struct page *page = bh->b_page;
	char *addr;
	__u32 checksum;

	addr = kmap_atomic(page);
	checksum = crc32_be(crc32_sum,
		(void *)(addr + offset_in_page(bh->b_data)), bh->b_size);
	kunmap_atomic(addr);

	return checksum;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
girish shilamkar
girish shilamkar
71
100.00%
1
100.00%
Total
71
100.00%
1
100.00%



static void write_tag_block(journal_t *j, journal_block_tag_t *tag,
				   unsigned long long block)
{
	tag->t_blocknr = cpu_to_be32(block & (u32)~0);
	if (jbd2_has_feature_64bit(j))
		tag->t_blocknr_high = cpu_to_be32((block >> 31) >> 1);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
zach brown
zach brown
46
82.14%
1
25.00%
darrick j. wong
darrick j. wong
7
12.50%
2
50.00%
mingming cao
mingming cao
3
5.36%
1
25.00%
Total
56
100.00%
4
100.00%



static void jbd2_block_tag_csum_set(journal_t *j, journal_block_tag_t *tag,
				    struct buffer_head *bh, __u32 sequence)
{
	journal_block_tag3_t *tag3 = (journal_block_tag3_t *)tag;
	struct page *page = bh->b_page;
	__u8 *addr;
	__u32 csum32;
	__be32 seq;

	if (!jbd2_journal_has_csum_v2or3(j))
		return;

	seq = cpu_to_be32(sequence);
	addr = kmap_atomic(page);
	csum32 = jbd2_chksum(j, j->j_csum_seed, (__u8 *)&seq, sizeof(seq));
	csum32 = jbd2_chksum(j, csum32, addr + offset_in_page(bh->b_data),
			     bh->b_size);
	kunmap_atomic(addr);

	if (jbd2_has_feature_csum3(j))
		tag3->t_checksum = cpu_to_be32(csum32);
	else
		tag->t_checksum = cpu_to_be16(csum32);
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
darrick j. wong
darrick j. wong
150
100.00%
5
100.00%
Total
150
100.00%
5
100.00%

/*
 * jbd2_journal_commit_transaction
 *
 * The primary function for committing a transaction to the log.  This
 * function is called by the journal thread to begin a complete commit.
 */


void jbd2_journal_commit_transaction(journal_t *journal)
{
	struct transaction_stats_s stats;
	transaction_t *commit_transaction;
	struct journal_head *jh;
	struct buffer_head *descriptor;
	struct buffer_head **wbuf = journal->j_wbuf;
	int bufs;
	int flags;
	int err;
	unsigned long long blocknr;
	ktime_t start_time;
	u64 commit_time;
	char *tagp = NULL;
	journal_block_tag_t *tag = NULL;
	int space_left = 0;
	int first_tag = 0;
	int tag_flag;
	int i;
	int tag_bytes = journal_tag_bytes(journal);
	struct buffer_head *cbh = NULL; /* For transactional checksums */
	__u32 crc32_sum = ~0;
	struct blk_plug plug;
	/* Tail of the journal */
	unsigned long first_block;
	tid_t first_tid;
	int update_tail;
	int csum_size = 0;
	LIST_HEAD(io_bufs);
	LIST_HEAD(log_bufs);

	if (jbd2_journal_has_csum_v2or3(journal))
		csum_size = sizeof(struct jbd2_journal_block_tail);

	/*
         * First job: lock down the current transaction and wait for
         * all outstanding updates to complete.
         */

	/* Do we need to erase the effects of a prior jbd2_journal_flush? */
	if (journal->j_flags & JBD2_FLUSHED) {
		jbd_debug(3, "super block updated\n");
		mutex_lock(&journal->j_checkpoint_mutex);
		/*
                 * We hold j_checkpoint_mutex so tail cannot change under us.
                 * We don't need any special data guarantees for writing sb
                 * since journal is empty and it is ok for write to be
                 * flushed only with transaction commit.
                 */
		jbd2_journal_update_sb_log_tail(journal,
						journal->j_tail_sequence,
						journal->j_tail,
						REQ_SYNC);
		mutex_unlock(&journal->j_checkpoint_mutex);
	} else {
		jbd_debug(3, "superblock not updated\n");
	}

	J_ASSERT(journal->j_running_transaction != NULL);
	J_ASSERT(journal->j_committing_transaction == NULL);

	commit_transaction = journal->j_running_transaction;

	trace_jbd2_start_commit(journal, commit_transaction);
	jbd_debug(1, "JBD2: starting commit of transaction %d\n",
			commit_transaction->t_tid);

	write_lock(&journal->j_state_lock);
	J_ASSERT(commit_transaction->t_state == T_RUNNING);
	commit_transaction->t_state = T_LOCKED;

	trace_jbd2_commit_locking(journal, commit_transaction);
	stats.run.rs_wait = commit_transaction->t_max_wait;
	stats.run.rs_request_delay = 0;
	stats.run.rs_locked = jiffies;
	if (commit_transaction->t_requested)
		stats.run.rs_request_delay =
			jbd2_time_diff(commit_transaction->t_requested,
				       stats.run.rs_locked);
	stats.run.rs_running = jbd2_time_diff(commit_transaction->t_start,
					      stats.run.rs_locked);

	spin_lock(&commit_transaction->t_handle_lock);
	while (atomic_read(&commit_transaction->t_updates)) {
		DEFINE_WAIT(wait);

		prepare_to_wait(&journal->j_wait_updates, &wait,
					TASK_UNINTERRUPTIBLE);
		if (atomic_read(&commit_transaction->t_updates)) {
			spin_unlock(&commit_transaction->t_handle_lock);
			write_unlock(&journal->j_state_lock);
			schedule();
			write_lock(&journal->j_state_lock);
			spin_lock(&commit_transaction->t_handle_lock);
		}
		finish_wait(&journal->j_wait_updates, &wait);
	}
	spin_unlock(&commit_transaction->t_handle_lock);

	J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <=
			journal->j_max_transaction_buffers);

	/*
         * First thing we are allowed to do is to discard any remaining
         * BJ_Reserved buffers.  Note, it is _not_ permissible to assume
         * that there are no such buffers: if a large filesystem
         * operation like a truncate needs to split itself over multiple
         * transactions, then it may try to do a jbd2_journal_restart() while
         * there are still BJ_Reserved buffers outstanding.  These must
         * be released cleanly from the current transaction.
         *
         * In this case, the filesystem must still reserve write access
         * again before modifying the buffer in the new transaction, but
         * we do not require it to remember exactly which old buffers it
         * has reserved.  This is consistent with the existing behaviour
         * that multiple jbd2_journal_get_write_access() calls to the same
         * buffer are perfectly permissible.
         */
	while (commit_transaction->t_reserved_list) {
		jh = commit_transaction->t_reserved_list;
		JBUFFER_TRACE(jh, "reserved, unused: refile");
		/*
                 * A jbd2_journal_get_undo_access()+jbd2_journal_release_buffer() may
                 * leave undo-committed data.
                 */
		if (jh->b_committed_data) {
			struct buffer_head *bh = jh2bh(jh);

			jbd_lock_bh_state(bh);
			jbd2_free(jh->b_committed_data, bh->b_size);
			jh->b_committed_data = NULL;
			jbd_unlock_bh_state(bh);
		}
		jbd2_journal_refile_buffer(journal, jh);
	}

	/*
         * Now try to drop any written-back buffers from the journal's
         * checkpoint lists.  We do this *before* commit because it potentially
         * frees some memory
         */
	spin_lock(&journal->j_list_lock);
	__jbd2_journal_clean_checkpoint_list(journal, false);
	spin_unlock(&journal->j_list_lock);

	jbd_debug(3, "JBD2: commit phase 1\n");

	/*
         * Clear revoked flag to reflect there is no revoked buffers
         * in the next transaction which is going to be started.
         */
	jbd2_clear_buffer_revoked_flags(journal);

	/*
         * Switch to a new revoke table.
         */
	jbd2_journal_switch_revoke_table(journal);

	/*
         * Reserved credits cannot be claimed anymore, free them
         */
	atomic_sub(atomic_read(&journal->j_reserved_credits),
		   &commit_transaction->t_outstanding_credits);

	trace_jbd2_commit_flushing(journal, commit_transaction);
	stats.run.rs_flushing = jiffies;
	stats.run.rs_locked = jbd2_time_diff(stats.run.rs_locked,
					     stats.run.rs_flushing);

	commit_transaction->t_state = T_FLUSH;
	journal->j_committing_transaction = commit_transaction;
	journal->j_running_transaction = NULL;
	start_time = ktime_get();
	commit_transaction->t_log_start = journal->j_head;
	wake_up(&journal->j_wait_transaction_locked);
	write_unlock(&journal->j_state_lock);

	jbd_debug(3, "JBD2: commit phase 2a\n");

	/*
         * Now start flushing things to disk, in the order they appear
         * on the transaction lists.  Data blocks go first.
         */
	err = journal_submit_data_buffers(journal, commit_transaction);
	if (err)
		jbd2_journal_abort(journal, err);

	blk_start_plug(&plug);
	jbd2_journal_write_revoke_records(commit_transaction, &log_bufs);

	jbd_debug(3, "JBD2: commit phase 2b\n");

	/*
         * Way to go: we have now written out all of the data for a
         * transaction!  Now comes the tricky part: we need to write out
         * metadata.  Loop over the transaction's entire buffer list:
         */
	write_lock(&journal->j_state_lock);
	commit_transaction->t_state = T_COMMIT;
	write_unlock(&journal->j_state_lock);

	trace_jbd2_commit_logging(journal, commit_transaction);
	stats.run.rs_logging = jiffies;
	stats.run.rs_flushing = jbd2_time_diff(stats.run.rs_flushing,
					       stats.run.rs_logging);
	stats.run.rs_blocks =
		atomic_read(&commit_transaction->t_outstanding_credits);
	stats.run.rs_blocks_logged = 0;

	J_ASSERT(commit_transaction->t_nr_buffers <=
		 atomic_read(&commit_transaction->t_outstanding_credits));

	err = 0;
	bufs = 0;
	descriptor = NULL;
	while (commit_transaction->t_buffers) {

		/* Find the next buffer to be journaled... */

		jh = commit_transaction->t_buffers;

		/* If we're in abort mode, we just un-journal the buffer and
                   release it. */

		if (is_journal_aborted(journal)) {
			clear_buffer_jbddirty(jh2bh(jh));
			JBUFFER_TRACE(jh, "journal is aborting: refile");
			jbd2_buffer_abort_trigger(jh,
						  jh->b_frozen_data ?
						  jh->b_frozen_triggers :
						  jh->b_triggers);
			jbd2_journal_refile_buffer(journal, jh);
			/* If that was the last one, we need to clean up
                         * any descriptor buffers which may have been
                         * already allocated, even if we are now
                         * aborting. */
			if (!commit_transaction->t_buffers)
				goto start_journal_io;
			continue;
		}

		/* Make sure we have a descriptor block in which to
                   record the metadata buffer. */

		if (!descriptor) {
			J_ASSERT (bufs == 0);

			jbd_debug(4, "JBD2: get descriptor\n");

			descriptor = jbd2_journal_get_descriptor_buffer(
							commit_transaction,
							JBD2_DESCRIPTOR_BLOCK);
			if (!descriptor) {
				jbd2_journal_abort(journal, -EIO);
				continue;
			}

			jbd_debug(4, "JBD2: got buffer %llu (%p)\n",
				(unsigned long long)descriptor->b_blocknr,
				descriptor->b_data);
			tagp = &descriptor->b_data[sizeof(journal_header_t)];
			space_left = descriptor->b_size -
						sizeof(journal_header_t);
			first_tag = 1;
			set_buffer_jwrite(descriptor);
			set_buffer_dirty(descriptor);
			wbuf[bufs++] = descriptor;

			/* Record it so that we can wait for IO
                           completion later */
			BUFFER_TRACE(descriptor, "ph3: file as descriptor");
			jbd2_file_log_bh(&log_bufs, descriptor);
		}

		/* Where is the buffer to be written? */

		err = jbd2_journal_next_log_block(journal, &blocknr);
		/* If the block mapping failed, just abandon the buffer
                   and repeat this loop: we'll fall into the
                   refile-on-abort condition above. */
		if (err) {
			jbd2_journal_abort(journal, err);
			continue;
		}

		/*
                 * start_this_handle() uses t_outstanding_credits to determine
                 * the free space in the log, but this counter is changed
                 * by jbd2_journal_next_log_block() also.
                 */
		atomic_dec(&commit_transaction->t_outstanding_credits);

		/* Bump b_count to prevent truncate from stumbling over
                   the shadowed buffer!  @@@ This can go if we ever get
                   rid of the shadow pairing of buffers. */
		atomic_inc(&jh2bh(jh)->b_count);

		/*
                 * Make a temporary IO buffer with which to write it out
                 * (this will requeue the metadata buffer to BJ_Shadow).
                 */
		set_bit(BH_JWrite, &jh2bh(jh)->b_state);
		JBUFFER_TRACE(jh, "ph3: write metadata");
		flags = jbd2_journal_write_metadata_buffer(commit_transaction,
						jh, &wbuf[bufs], blocknr);
		if (flags < 0) {
			jbd2_journal_abort(journal, flags);
			continue;
		}
		jbd2_file_log_bh(&io_bufs, wbuf[bufs]);

		/* Record the new block's tag in the current descriptor
                   buffer */

		tag_flag = 0;
		if (flags & 1)
			tag_flag |= JBD2_FLAG_ESCAPE;
		if (!first_tag)
			tag_flag |= JBD2_FLAG_SAME_UUID;

		tag = (journal_block_tag_t *) tagp;
		write_tag_block(journal, tag, jh2bh(jh)->b_blocknr);
		tag->t_flags = cpu_to_be16(tag_flag);
		jbd2_block_tag_csum_set(journal, tag, wbuf[bufs],
					commit_transaction->t_tid);
		tagp += tag_bytes;
		space_left -= tag_bytes;
		bufs++;

		if (first_tag) {
			memcpy (tagp, journal->j_uuid, 16);
			tagp += 16;
			space_left -= 16;
			first_tag = 0;
		}

		/* If there's no more to do, or if the descriptor is full,
                   let the IO rip! */

		if (bufs == journal->j_wbufsize ||
		    commit_transaction->t_buffers == NULL ||
		    space_left < tag_bytes + 16 + csum_size) {

			jbd_debug(4, "JBD2: Submit %d IOs\n", bufs);

			/* Write an end-of-descriptor marker before
                           submitting the IOs.  "tag" still points to
                           the last tag we set up. */

			tag->t_flags |= cpu_to_be16(JBD2_FLAG_LAST_TAG);

			jbd2_descriptor_block_csum_set(journal, descriptor);
start_journal_io:
			for (i = 0; i < bufs; i++) {
				struct buffer_head *bh = wbuf[i];
				/*
                                 * Compute checksum.
                                 */
				if (jbd2_has_feature_checksum(journal)) {
					crc32_sum =
					    jbd2_checksum_data(crc32_sum, bh);
				}

				lock_buffer(bh);
				clear_buffer_dirty(bh);
				set_buffer_uptodate(bh);
				bh->b_end_io = journal_end_buffer_io_sync;
				submit_bh(REQ_OP_WRITE, REQ_SYNC, bh);
			}
			cond_resched();
			stats.run.rs_blocks_logged += bufs;

			/* Force a new descriptor to be generated next
                           time round the loop. */
			descriptor = NULL;
			bufs = 0;
		}
	}

	err = journal_finish_inode_data_buffers(journal, commit_transaction);
	if (err) {
		printk(KERN_WARNING
			"JBD2: Detected IO errors while flushing file data "
		       "on %s\n", journal->j_devname);
		if (journal->j_flags & JBD2_ABORT_ON_SYNCDATA_ERR)
			jbd2_journal_abort(journal, err);
		err = 0;
	}

	/*
         * Get current oldest transaction in the log before we issue flush
         * to the filesystem device. After the flush we can be sure that
         * blocks of all older transactions are checkpointed to persistent
         * storage and we will be safe to update journal start in the
         * superblock with the numbers we get here.
         */
	update_tail =
		jbd2_journal_get_log_tail(journal, &first_tid, &first_block);

	write_lock(&journal->j_state_lock);
	if (update_tail) {
		long freed = first_block - journal->j_tail;

		if (first_block < journal->j_tail)
			freed += journal->j_last - journal->j_first;
		/* Update tail only if we free significant amount of space */
		if (freed < journal->j_maxlen / 4)
			update_tail = 0;
	}
	J_ASSERT(commit_transaction->t_state == T_COMMIT);
	commit_transaction->t_state = T_COMMIT_DFLUSH;
	write_unlock(&journal->j_state_lock);

	/* 
         * If the journal is not located on the file system device,
         * then we must flush the file system device before we issue
         * the commit record
         */
	if (commit_transaction->t_need_data_flush &&
	    (journal->j_fs_dev != journal->j_dev) &&
	    (journal->j_flags & JBD2_BARRIER))
		blkdev_issue_flush(journal->j_fs_dev, GFP_NOFS, NULL);

	/* Done it all: now write the commit record asynchronously. */
	if (jbd2_has_fea