Ejemplo n.º 1
0
static void
__journal_try_to_free_buffer(journal_t *journal, struct buffer_head *bh)
{
	struct journal_head *jh;

	jh = bh2jh(bh);

	if (buffer_locked(bh) || buffer_dirty(bh))
		goto out;

	if (jh->b_next_transaction != NULL)
		goto out;

	spin_lock(&journal->j_list_lock);
	if (jh->b_transaction != NULL && jh->b_cp_transaction == NULL) {
		if (jh->b_jlist == BJ_SyncData || jh->b_jlist == BJ_Locked) {
			/* A written-back ordered data buffer */
			JBUFFER_TRACE(jh, "release data");
			__journal_unfile_buffer(jh);
			journal_remove_journal_head(bh);
			__brelse(bh);
		}
	} else if (jh->b_cp_transaction != NULL && jh->b_transaction == NULL) {
		/* written-back checkpointed metadata buffer */
		if (jh->b_jlist == BJ_None) {
			JBUFFER_TRACE(jh, "remove from checkpoint list");
			__journal_remove_checkpoint(jh);
			journal_remove_journal_head(bh);
			__brelse(bh);
		}
	}
	spin_unlock(&journal->j_list_lock);
out:
	return;
}
Ejemplo n.º 2
0
Archivo: buffer.c Proyecto: foolsh/elks
static void sync_buffers(kdev_t dev, int wait)
{
    register struct buffer_head *bh;

    for (bh = bh_chain; bh != NULL; bh = bh->b_next) {
	if (dev && bh->b_dev != dev)
	    continue;
	/*
	 *      Skip clean buffers.
	 */
	if (buffer_clean(bh))
	    continue;
	/*
	 *      Locked buffers..
	 *
	 *      Buffer is locked; skip it unless wait is requested
	 *      AND pass > 0.
	 */
	if (buffer_locked(bh) && wait)
	    continue;
	else
	    wait_on_buffer(bh);
	/*
	 *      Do the stuff
	 */
	bh->b_count++;
	ll_rw_blk(WRITE, bh);
	bh->b_count--;
    }
    return;
}
Ejemplo n.º 3
0
static void sprintf_buffer_head (char * buf, struct buffer_head * bh) 
{
  sprintf (buf, "dev %s, size %d, blocknr %ld, count %d, list %d, state 0x%lx, page %p, (%s, %s, %s)",
	   kdevname (bh->b_dev), bh->b_size, bh->b_blocknr, atomic_read (&(bh->b_count)), bh->b_list,
	   bh->b_state, bh->b_page,
	   buffer_uptodate (bh) ? "UPTODATE" : "!UPTODATE",
	   buffer_dirty (bh) ? "DIRTY" : "CLEAN",
	   buffer_locked (bh) ? "LOCKED" : "UNLOCKED");
}
Ejemplo n.º 4
0
/*
 * Try to flush one buffer from the checkpoint list to disk.
 *
 * Return 1 if something happened which requires us to abort the current
 * scan of the checkpoint list.  
 *
 * Called with j_list_lock held and drops it if 1 is returned
 * Called under jbd_lock_bh_state(jh2bh(jh)), and drops it
 */
static int __process_buffer(journal_t *journal, struct journal_head *jh,
			struct buffer_head **bhs, int *batch_count)
{
	struct buffer_head *bh = jh2bh(jh);
	int ret = 0;

	if (buffer_locked(bh)) {
		atomic_inc(&bh->b_count);
		spin_unlock(&journal->j_list_lock);
		jbd_unlock_bh_state(bh);
		wait_on_buffer(bh);
		/* the journal_head may have gone by now */
		BUFFER_TRACE(bh, "brelse");
		__brelse(bh);
		ret = 1;
	} else if (jh->b_transaction != NULL) {
		transaction_t *t = jh->b_transaction;
		tid_t tid = t->t_tid;

		spin_unlock(&journal->j_list_lock);
		jbd_unlock_bh_state(bh);
		log_start_commit(journal, tid);
		log_wait_commit(journal, tid);
		ret = 1;
	} else if (!buffer_dirty(bh)) {
		J_ASSERT_JH(jh, !buffer_jbddirty(bh));
		BUFFER_TRACE(bh, "remove from checkpoint");
		__journal_remove_checkpoint(jh);
		spin_unlock(&journal->j_list_lock);
		jbd_unlock_bh_state(bh);
		journal_remove_journal_head(bh);
		__brelse(bh);
		ret = 1;
	} else {
		/*
		 * Important: we are about to write the buffer, and
		 * possibly block, while still holding the journal lock.
		 * We cannot afford to let the transaction logic start
		 * messing around with this buffer before we write it to
		 * disk, as that would break recoverability.  
		 */
		BUFFER_TRACE(bh, "queue");
		get_bh(bh);
		J_ASSERT_BH(bh, !buffer_jwrite(bh));
		set_buffer_jwrite(bh);
		bhs[*batch_count] = bh;
		__buffer_relink_io(jh);
		jbd_unlock_bh_state(bh);
		(*batch_count)++;
		if (*batch_count == NR_BATCH) {
			spin_unlock(&journal->j_list_lock);
			__flush_batch(journal, bhs, batch_count);
			ret = 1;
		}
	}
	return ret;
}
Ejemplo n.º 5
0
static void lock_journal_bh_wait(struct buffer_head *bh, journal_t *journal)
{
repeat:
	wait_on_buffer(bh);
	lock_journal(journal);
	if (buffer_locked(bh)) {
		unlock_journal(journal);
		goto repeat;
	}
}
Ejemplo n.º 6
0
static int do_readahead(journal_t *journal, unsigned int start)
{
	int err;
	unsigned int max, nbufs, next, blocknr;
	struct buffer_head *bh;
	
	struct buffer_head * bufs[MAXBUF];
	
	/* Do up to 128K of readahead */
	max = start + (128 * 1024 / journal->j_blocksize);
	if (max > journal->j_maxlen)
		max = journal->j_maxlen;

	/* Do the readahead itself.  We'll submit MAXBUF buffer_heads at
	 * a time to the block device IO layer. */
	
	nbufs = 0;
	
	for (next = start; next < max; next++) {
		blocknr = next;
		if (journal->j_inode)
			blocknr = bmap(journal->j_inode, next);
		if (!blocknr) {
			printk (KERN_ERR "JFS: bad block at offset %u\n",
				next);
			err = -EIO;
			goto failed;
		}
		
		bh = getblk(journal->j_dev, blocknr, journal->j_blocksize);
		if (!bh) {
			err = -ENOMEM;
			goto failed;
		}

		if (!buffer_uptodate(bh) && !buffer_locked(bh)) {
			bufs[nbufs++] = bh;
			if (nbufs == MAXBUF) {
				ll_rw_block(READ, nbufs, bufs);
				brelse_array(bufs, nbufs);
				nbufs = 0;
			}
		} else
			brelse(bh);
	}

	if (nbufs)
		ll_rw_block(READ, nbufs, bufs);
	err = 0;
	
failed:	
	if (nbufs) 
		brelse_array(bufs, nbufs);
	return err;
}
Ejemplo n.º 7
0
Archivo: buffer.c Proyecto: foolsh/elks
void wait_on_buffer(register struct buffer_head *bh)
{
    if (buffer_locked(bh)) {

	bh->b_count++;

	wait_set(&bh->b_wait);

	goto chk_buf;
	do {
	    current->state = TASK_UNINTERRUPTIBLE;
	    schedule();
	    current->state = TASK_RUNNING;
    chk_buf:
	    ;
	} while(buffer_locked(bh));

	wait_clear(&bh->b_wait);
	bh->b_count--;
    }
}
Ejemplo n.º 8
0
/*
 * Try to release a checkpointed buffer from its transaction.
 * Returns 1 if we released it and 2 if we also released the
 * whole transaction.
 *
 * Requires j_list_lock
 */
static int __try_to_free_cp_buf(struct journal_head *jh)
{
	int ret = 0;
	struct buffer_head *bh = jh2bh(jh);

	if (jh->b_transaction == NULL && !buffer_locked(bh) &&
	    !buffer_dirty(bh) && !buffer_write_io_error(bh)) {
		JBUFFER_TRACE(jh, "remove from checkpoint list");
		ret = __jbd2_journal_remove_checkpoint(jh) + 1;
	}
	return ret;
}
Ejemplo n.º 9
0
static int __process_buffer(journal_t *journal, struct journal_head *jh,
                            int *batch_count, transaction_t *transaction)
{
    struct buffer_head *bh = jh2bh(jh);
    int ret = 0;

    if (buffer_locked(bh)) {
        get_bh(bh);
        spin_unlock(&journal->j_list_lock);
        wait_on_buffer(bh);

        BUFFER_TRACE(bh, "brelse");
        __brelse(bh);
        ret = 1;
    } else if (jh->b_transaction != NULL) {
        transaction_t *t = jh->b_transaction;
        tid_t tid = t->t_tid;

        transaction->t_chp_stats.cs_forced_to_close++;
        spin_unlock(&journal->j_list_lock);
        if (unlikely(journal->j_flags & JBD2_UNMOUNT))
            printk(KERN_ERR "JBD2: %s: "
                   "Waiting for Godot: block %llu\n",
                   journal->j_devname,
                   (unsigned long long) bh->b_blocknr);
        jbd2_log_start_commit(journal, tid);
        jbd2_log_wait_commit(journal, tid);
        ret = 1;
    } else if (!buffer_dirty(bh)) {
        ret = 1;
        if (unlikely(buffer_write_io_error(bh)))
            ret = -EIO;
        get_bh(bh);
        BUFFER_TRACE(bh, "remove from checkpoint");
        __jbd2_journal_remove_checkpoint(jh);
        spin_unlock(&journal->j_list_lock);
        __brelse(bh);
    } else {
        BUFFER_TRACE(bh, "queue");
        get_bh(bh);
        J_ASSERT_BH(bh, !buffer_jwrite(bh));
        journal->j_chkpt_bhs[*batch_count] = bh;
        __buffer_relink_io(jh);
        transaction->t_chp_stats.cs_written++;
        (*batch_count)++;
        if (*batch_count == JBD2_NR_BATCH) {
            spin_unlock(&journal->j_list_lock);
            __flush_batch(journal, batch_count);
            ret = 1;
        }
    }
    return ret;
}
Ejemplo n.º 10
0
static int do_readahead(journal_t *journal, unsigned int start)
{
	int err;
	unsigned int max, nbufs, next;
	unsigned long long blocknr;
	struct buffer_head *bh;

	struct buffer_head * bufs[MAXBUF];

	
	max = start + (128 * 1024 / journal->j_blocksize);
	if (max > journal->j_maxlen)
		max = journal->j_maxlen;


	nbufs = 0;

	for (next = start; next < max; next++) {
		err = jbd2_journal_bmap(journal, next, &blocknr);

		if (err) {
			printk(KERN_ERR "JBD2: bad block at offset %u\n",
				next);
			goto failed;
		}

		bh = __getblk(journal->j_dev, blocknr, journal->j_blocksize);
		if (!bh) {
			err = -ENOMEM;
			goto failed;
		}

		if (!buffer_uptodate(bh) && !buffer_locked(bh)) {
			bufs[nbufs++] = bh;
			if (nbufs == MAXBUF) {
				ll_rw_block(READ, nbufs, bufs);
				journal_brelse_array(bufs, nbufs);
				nbufs = 0;
			}
		} else
			brelse(bh);
	}

	if (nbufs)
		ll_rw_block(READ, nbufs, bufs);
	err = 0;

failed:
	if (nbufs)
		journal_brelse_array(bufs, nbufs);
	return err;
}
Ejemplo n.º 11
0
Archivo: prints.c Proyecto: 7799/linux
static void sprintf_buffer_head(char *buf, struct buffer_head *bh)
{
	char b[BDEVNAME_SIZE];

	sprintf(buf,
		"dev %s, size %zd, blocknr %llu, count %d, state 0x%lx, page %p, (%s, %s, %s)",
		bdevname(bh->b_bdev, b), bh->b_size,
		(unsigned long long)bh->b_blocknr, atomic_read(&(bh->b_count)),
		bh->b_state, bh->b_page,
		buffer_uptodate(bh) ? "UPTODATE" : "!UPTODATE",
		buffer_dirty(bh) ? "DIRTY" : "CLEAN",
		buffer_locked(bh) ? "LOCKED" : "UNLOCKED");
}
Ejemplo n.º 12
0
Archivo: mdt.c Proyecto: AK101111/linux
static int nilfs_mdt_read_block(struct inode *inode, unsigned long block,
				int readahead, struct buffer_head **out_bh)
{
	struct buffer_head *first_bh, *bh;
	unsigned long blkoff;
	int i, nr_ra_blocks = NILFS_MDT_MAX_RA_BLOCKS;
	int err;

	err = nilfs_mdt_submit_block(inode, block, REQ_OP_READ, 0, &first_bh);
	if (err == -EEXIST) /* internal code */
		goto out;

	if (unlikely(err))
		goto failed;

	if (readahead) {
		blkoff = block + 1;
		for (i = 0; i < nr_ra_blocks; i++, blkoff++) {
			err = nilfs_mdt_submit_block(inode, blkoff, REQ_OP_READ,
						     REQ_RAHEAD, &bh);
			if (likely(!err || err == -EEXIST))
				brelse(bh);
			else if (err != -EBUSY)
				break;
				/* abort readahead if bmap lookup failed */
			if (!buffer_locked(first_bh))
				goto out_no_wait;
		}
	}

	wait_on_buffer(first_bh);

 out_no_wait:
	err = -EIO;
	if (!buffer_uptodate(first_bh)) {
		nilfs_msg(inode->i_sb, KERN_ERR,
			  "I/O error reading meta-data file (ino=%lu, block-offset=%lu)",
			  inode->i_ino, block);
		goto failed_bh;
	}
 out:
	*out_bh = first_bh;
	return 0;

 failed_bh:
	brelse(first_bh);
 failed:
	return err;
}
Ejemplo n.º 13
0
int jbd2_journal_get_create_access(handle_t *handle, struct buffer_head *bh)
{
	transaction_t *transaction = handle->h_transaction;
	journal_t *journal = transaction->t_journal;
	struct journal_head *jh = jbd2_journal_add_journal_head(bh);
	int err;

	jbd_debug(5, "journal_head %p\n", jh);
	err = -EROFS;
	if (is_handle_aborted(handle))
		goto out;
	err = 0;

	JBUFFER_TRACE(jh, "entry");
	jbd_lock_bh_state(bh);
	spin_lock(&journal->j_list_lock);
	J_ASSERT_JH(jh, (jh->b_transaction == transaction ||
		jh->b_transaction == NULL ||
		(jh->b_transaction == journal->j_committing_transaction &&
			  jh->b_jlist == BJ_Forget)));

	J_ASSERT_JH(jh, jh->b_next_transaction == NULL);
	J_ASSERT_JH(jh, buffer_locked(jh2bh(jh)));

	if (jh->b_transaction == NULL) {
		clear_buffer_dirty(jh2bh(jh));
		
		jh->b_modified = 0;

		JBUFFER_TRACE(jh, "file as BJ_Reserved");
		__jbd2_journal_file_buffer(jh, transaction, BJ_Reserved);
	} else if (jh->b_transaction == journal->j_committing_transaction) {
		
		jh->b_modified = 0;

		JBUFFER_TRACE(jh, "set next transaction");
		jh->b_next_transaction = transaction;
	}
	spin_unlock(&journal->j_list_lock);
	jbd_unlock_bh_state(bh);

	JBUFFER_TRACE(jh, "cancelling revoke");
	jbd2_journal_cancel_revoke(handle, jh);
out:
	jbd2_journal_put_journal_head(jh);
	return err;
}
Ejemplo n.º 14
0
/*
 * Try to release a checkpointed buffer from its transaction.
 * Returns 1 if we released it and 2 if we also released the
 * whole transaction.
 *
 * Requires j_list_lock
 * Called under jbd_lock_bh_state(jh2bh(jh)), and drops it
 */
static int __try_to_free_cp_buf(struct journal_head *jh)
{
	int ret = 0;
	struct buffer_head *bh = jh2bh(jh);

	if (jh->b_jlist == BJ_None && !buffer_locked(bh) && !buffer_dirty(bh)) {
		JBUFFER_TRACE(jh, "remove from checkpoint list");
		ret = __jbd2_journal_remove_checkpoint(jh) + 1;
		jbd_unlock_bh_state(bh);
		jbd2_journal_remove_journal_head(bh);
		BUFFER_TRACE(bh, "release");
		__brelse(bh);
	} else {
		jbd_unlock_bh_state(bh);
	}
	return ret;
}
Ejemplo n.º 15
0
/*
 * Try to release a checkpointed buffer from its transaction.
 * Returns 1 if we released it.
 * Requires journal_datalist_lock
 */
static int __try_to_free_cp_buf(struct journal_head *jh)
{
	int ret = 0;
	struct buffer_head *bh = jh2bh(jh);

	if (jh->b_jlist == BJ_None && !buffer_locked(bh) && !buffer_dirty(bh)) {
		JBUFFER_TRACE(jh, "remove from checkpoint list");
		__journal_remove_checkpoint(jh);
		__journal_remove_journal_head(bh);
		BUFFER_TRACE(bh, "release");
		/* BUF_LOCKED -> BUF_CLEAN (fwiw) */
		refile_buffer(bh);
		__brelse(bh);
		ret = 1;
	}
	return ret;
}
Ejemplo n.º 16
0
/*
 * Try to release a checkpointed buffer from its transaction.
 * Returns 1 if we released it and 2 if we also released the
 * whole transaction.
 *
 * Requires j_list_lock
 */
static int __try_to_free_cp_buf(struct journal_head *jh)
{
	int ret = 0;
	struct buffer_head *bh = jh2bh(jh);

	if (jh->b_transaction == NULL && !buffer_locked(bh) &&
	    !buffer_dirty(bh) && !buffer_write_io_error(bh)) {
		/*
		 * Get our reference so that bh cannot be freed before
		 * we unlock it
		 */
		get_bh(bh);
		JBUFFER_TRACE(jh, "remove from checkpoint list");
		ret = __jbd2_journal_remove_checkpoint(jh) + 1;
		BUFFER_TRACE(bh, "release");
		__brelse(bh);
	}
	return ret;
}
Ejemplo n.º 17
0
/*
 * Try to flush one buffer from the checkpoint list to disk.
 *
 * Return 1 if something happened which requires us to abort the current
 * scan of the checkpoint list.  
 *
 * Called with j_list_lock held.
 * Called under jbd_lock_bh_state(jh2bh(jh)), and drops it
 */
static int __flush_buffer(journal_t *journal, struct journal_head *jh,
			struct buffer_head **bhs, int *batch_count,
			int *drop_count)
{
	struct buffer_head *bh = jh2bh(jh);
	int ret = 0;

	if (buffer_dirty(bh) && !buffer_locked(bh) && jh->b_jlist == BJ_None) {
		J_ASSERT_JH(jh, jh->b_transaction == NULL);

		/*
		 * Important: we are about to write the buffer, and
		 * possibly block, while still holding the journal lock.
		 * We cannot afford to let the transaction logic start
		 * messing around with this buffer before we write it to
		 * disk, as that would break recoverability.  
		 */
		BUFFER_TRACE(bh, "queue");
		get_bh(bh);
		J_ASSERT_BH(bh, !buffer_jwrite(bh));
		set_buffer_jwrite(bh);
		bhs[*batch_count] = bh;
		jbd_unlock_bh_state(bh);
		(*batch_count)++;
		if (*batch_count == NR_BATCH) {
			__flush_batch(journal, bhs, batch_count);
			ret = 1;
		}
	} else {
		int last_buffer = 0;
		if (jh->b_cpnext == jh) {
			/* We may be about to drop the transaction.  Tell the
			 * caller that the lists have changed.
			 */
			last_buffer = 1;
		}
		if (__try_to_free_cp_buf(jh)) {
			(*drop_count)++;
			ret = last_buffer;
		}
	}
	return ret;
}
Ejemplo n.º 18
0
/*
 * Clean up transaction's list of buffers submitted for io.
 * We wait for any pending IO to complete and remove any clean
 * buffers. Note that we take the buffers in the opposite ordering
 * from the one in which they were submitted for IO.
 *
 * Called with j_list_lock held.
 */
static void __wait_cp_io(journal_t *journal, transaction_t *transaction)
{
	struct journal_head *jh;
	struct buffer_head *bh;
	tid_t this_tid;
	int released = 0;

	this_tid = transaction->t_tid;
restart:
	/* Did somebody clean up the transaction in the meanwhile? */
	if (journal->j_checkpoint_transactions != transaction ||
			transaction->t_tid != this_tid)
		return;
	while (!released && transaction->t_checkpoint_io_list) {
		jh = transaction->t_checkpoint_io_list;
		bh = jh2bh(jh);
		if (!jbd_trylock_bh_state(bh)) {
			jbd_sync_bh(journal, bh);
			spin_lock(&journal->j_list_lock);
			goto restart;
		}
		if (buffer_locked(bh)) {
			atomic_inc(&bh->b_count);
			spin_unlock(&journal->j_list_lock);
			jbd_unlock_bh_state(bh);
			wait_on_buffer(bh);
			/* the journal_head may have gone by now */
			BUFFER_TRACE(bh, "brelse");
			__brelse(bh);
			spin_lock(&journal->j_list_lock);
			goto restart;
		}
		/*
		 * Now in whatever state the buffer currently is, we know that
		 * it has been written out and so we can drop it from the list
		 */
		released = __journal_remove_checkpoint(jh);
		jbd_unlock_bh_state(bh);
		journal_remove_journal_head(bh);
		__brelse(bh);
	}
}
Ejemplo n.º 19
0
static int nilfs_mdt_read_block(struct inode *inode, unsigned long block,
				struct buffer_head **out_bh)
{
	struct buffer_head *first_bh, *bh;
	unsigned long blkoff;
	int i, nr_ra_blocks = NILFS_MDT_MAX_RA_BLOCKS;
	int err;

	err = nilfs_mdt_submit_block(inode, block, READ, &first_bh);
	if (err == -EEXIST) /* internal code */
		goto out;

	if (unlikely(err))
		goto failed;

	blkoff = block + 1;
	for (i = 0; i < nr_ra_blocks; i++, blkoff++) {
		err = nilfs_mdt_submit_block(inode, blkoff, READA, &bh);
		if (likely(!err || err == -EEXIST))
			brelse(bh);
		else if (err != -EBUSY)
			break; /* abort readahead if bmap lookup failed */

		if (!buffer_locked(first_bh))
			goto out_no_wait;
	}

	wait_on_buffer(first_bh);

 out_no_wait:
	err = -EIO;
	if (!buffer_uptodate(first_bh))
		goto failed_bh;
 out:
	*out_bh = first_bh;
	return 0;

 failed_bh:
	brelse(first_bh);
 failed:
	return err;
}
Ejemplo n.º 20
0
static int __wait_cp_io(journal_t *journal, transaction_t *transaction)
{
	struct journal_head *jh;
	struct buffer_head *bh;
	tid_t this_tid;
	int released = 0;
	int ret = 0;

	this_tid = transaction->t_tid;
restart:
	/*                                                         */
	if (journal->j_checkpoint_transactions != transaction ||
			transaction->t_tid != this_tid)
		return ret;
	while (!released && transaction->t_checkpoint_io_list) {
		jh = transaction->t_checkpoint_io_list;
		bh = jh2bh(jh);
		get_bh(bh);
		if (buffer_locked(bh)) {
			spin_unlock(&journal->j_list_lock);
			wait_on_buffer(bh);
			/*                                       */
			BUFFER_TRACE(bh, "brelse");
			__brelse(bh);
			spin_lock(&journal->j_list_lock);
			goto restart;
		}
		if (unlikely(buffer_write_io_error(bh)))
			ret = -EIO;

		/*
                                                                
                                                                
   */
		released = __jbd2_journal_remove_checkpoint(jh);
		__brelse(bh);
	}

	return ret;
}
Ejemplo n.º 21
0
/*
 * Clean up transaction's list of buffers submitted for io.
 * We wait for any pending IO to complete and remove any clean
 * buffers. Note that we take the buffers in the opposite ordering
 * from the one in which they were submitted for IO.
 *
 * Return 0 on success, and return <0 if some buffers have failed
 * to be written out.
 *
 * Called with j_list_lock held.
 */
static int __wait_cp_io(journal_t *journal, transaction_t *transaction)
{
	struct journal_head *jh;
	struct buffer_head *bh;
	tid_t this_tid;
	int released = 0;
	int ret = 0;

	this_tid = transaction->t_tid;
restart:
	/* Did somebody clean up the transaction in the meanwhile? */
	if (journal->j_checkpoint_transactions != transaction ||
			transaction->t_tid != this_tid)
		return ret;
	while (!released && transaction->t_checkpoint_io_list) {
		jh = transaction->t_checkpoint_io_list;
		bh = jh2bh(jh);
		get_bh(bh);
		if (buffer_locked(bh)) {
			spin_unlock(&journal->j_list_lock);
			wait_on_buffer(bh);
			/* the journal_head may have gone by now */
			BUFFER_TRACE(bh, "brelse");
			__brelse(bh);
			spin_lock(&journal->j_list_lock);
			goto restart;
		}
		if (unlikely(buffer_write_io_error(bh)))
			ret = -EIO;

		/*
		 * Now in whatever state the buffer currently is, we know that
		 * it has been written out and so we can drop it from the list
		 */
		released = __jbd2_journal_remove_checkpoint(jh);
		__brelse(bh);
	}

	return ret;
}
Ejemplo n.º 22
0
static void do_stram_request( void )
{
	unsigned long start, len;

	while( CURRENT ) {
		if (MAJOR(CURRENT->rq_dev) != MAJOR_NR)
			panic("stram: request list destroyed");
		if (CURRENT->bh) {
			if (!buffer_locked(CURRENT->bh))
				panic("stram: block not locked");
		}
		
		start = swap_start + (CURRENT->sector << 9);
		len   = CURRENT->current_nr_sectors << 9;
		if ((start + len) > swap_end) {
			printk( KERN_ERR "stram: bad access beyond end of device: "
					"block=%ld, count=%ld\n",
					CURRENT->sector,
					CURRENT->current_nr_sectors );
			end_request( 0 );
			continue;
		}

		if (CURRENT->cmd == READ) {
			memcpy( CURRENT->buffer, (char *)start, len );
#ifdef DO_PROC
			stat_swap_read += N_PAGES(len);
#endif
		}
		else {
			memcpy( (char *)start, CURRENT->buffer, len );
#ifdef DO_PROC
			stat_swap_write += N_PAGES(len);
#endif
		}
		end_request( 1 );
	}
}
Ejemplo n.º 23
0
static void
__journal_try_to_free_buffer(journal_t *journal, struct buffer_head *bh)
{
	struct journal_head *jh;

	jh = bh2jh(bh);

	if (buffer_locked(bh) || buffer_dirty(bh))
		goto out;

	if (jh->b_next_transaction != NULL)
		goto out;

	spin_lock(&journal->j_list_lock);
	if (jh->b_cp_transaction != NULL && jh->b_transaction == NULL) {
		
		JBUFFER_TRACE(jh, "remove from checkpoint list");
		__jbd2_journal_remove_checkpoint(jh);
	}
	spin_unlock(&journal->j_list_lock);
out:
	return;
}
Ejemplo n.º 24
0
int journal_get_create_access (handle_t *handle, struct buffer_head *bh) 
{
	transaction_t *transaction = handle->h_transaction;
	journal_t *journal = transaction->t_journal;

	jfs_debug(5, "buffer_head %p\n", bh);
	lock_journal(journal);

	/* The buffer may already belong to this transaction due to
           pre-zeroing in the filesystem's new_block code */
	J_ASSERT (bh->b_transaction == transaction || bh->b_transaction == NULL);
	J_ASSERT (bh->b_next_transaction == NULL);
	J_ASSERT (buffer_locked(bh));

	J_ASSERT(handle->h_buffer_credits > 0);
	handle->h_buffer_credits--;
	
	bh->b_transaction = transaction;
	journal_file_buffer(bh, transaction, BJ_Reserved);
	refile_buffer(bh);
	unlock_journal(journal);
	return 0;
}
Ejemplo n.º 25
0
static void
do_mcd_request(void)
{
#ifdef TEST2
  printk(" do_mcd_request(%ld+%ld)\n", CURRENT -> sector, CURRENT -> nr_sectors);
#endif
  mcd_transfer_is_active = 1;
  while (CURRENT_VALID) {
    if (CURRENT->bh) {
      if (!buffer_locked(CURRENT->bh))
	panic(DEVICE_NAME ": block not locked");
    }
    mcd_transfer();
    if (CURRENT -> nr_sectors == 0) {
      end_request(1);
    } else {
      mcd_buf_out = -1;		/* Want to read a block not in buffer */
      if (mcd_state == MCD_S_IDLE) {
	if (!tocUpToDate) {
	  if (updateToc() < 0) {
	    while (CURRENT_VALID)
	      end_request(0);
	    break;
	  }
	}
	mcd_state = MCD_S_START;
	McdTries = 5;
	SET_TIMER(mcd_poll, 1);
      }
      break;
    }
  }
  mcd_transfer_is_active = 0;
#ifdef TEST2
  printk(" do_mcd_request ends\n");
#endif
}
Ejemplo n.º 26
0
/* it searches for a window of zero bits with given minimum and maximum lengths in one bitmap
 * block; */
static int scan_bitmap_block (struct reiserfs_transaction_handle *th,
                              int bmap_n, int *beg, int boundary, int min, int max, int unfm)
{
    struct super_block *s = th->t_super;
    struct reiserfs_bitmap_info *bi=&SB_AP_BITMAP(s)[bmap_n];
    int end, next;
    int org = *beg;

    BUG_ON (!th->t_trans_id);

    RFALSE(bmap_n >= SB_BMAP_NR (s), "Bitmap %d is out of range (0..%d)",bmap_n, SB_BMAP_NR (s) - 1);
    PROC_INFO_INC( s, scan_bitmap.bmap );
    /* this is unclear and lacks comments, explain how journal bitmaps
       work here for the reader.  Convey a sense of the design here. What
       is a window? */
    /* - I mean `a window of zero bits' as in description of this function - Zam. */

    if ( !bi ) {
        reiserfs_warning (s, "NULL bitmap info pointer for bitmap %d", bmap_n);
        return 0;
    }
    if (buffer_locked (bi->bh)) {
        PROC_INFO_INC( s, scan_bitmap.wait );
        __wait_on_buffer (bi->bh);
    }

    while (1) {
cont:
        if (bi->free_count < min)
            return 0; // No free blocks in this bitmap

        /* search for a first zero bit -- beggining of a window */
        *beg = reiserfs_find_next_zero_le_bit
               ((unsigned long*)(bi->bh->b_data), boundary, *beg);

        if (*beg + min > boundary) {
            /* search for a zero bit fails or the rest of bitmap block
            			      * cannot contain a zero window of minimum size */
            return 0;
        }

        if (unfm && is_block_in_journal(s,bmap_n, *beg, beg))
            continue;
        /* first zero bit found; we check next bits */
        for (end = *beg + 1;; end ++) {
            if (end >= *beg + max || end >= boundary || reiserfs_test_le_bit (end, bi->bh->b_data)) {
                next = end;
                break;
            }
            /* finding the other end of zero bit window requires looking into journal structures (in
             * case of searching for free blocks for unformatted nodes) */
            if (unfm && is_block_in_journal(s, bmap_n, end, &next))
                break;
        }

        /* now (*beg) points to beginning of zero bits window,
         * (end) points to one bit after the window end */
        if (end - *beg >= min) { /* it seems we have found window of proper size */
            int i;
            reiserfs_prepare_for_journal (s, bi->bh, 1);
            /* try to set all blocks used checking are they still free */
            for (i = *beg; i < end; i++) {
                /* It seems that we should not check in journal again. */
                if (reiserfs_test_and_set_le_bit (i, bi->bh->b_data)) {
                    /* bit was set by another process
                     * while we slept in prepare_for_journal() */
                    PROC_INFO_INC( s, scan_bitmap.stolen );
                    if (i >= *beg + min)	{
                        /* we can continue with smaller set of allocated blocks,
                        		   * if length of this set is more or equal to `min' */
                        end = i;
                        break;
                    }
                    /* otherwise we clear all bit were set ... */
                    while (--i >= *beg)
                        reiserfs_test_and_clear_le_bit (i, bi->bh->b_data);
                    reiserfs_restore_prepared_buffer (s, bi->bh);
                    *beg = org;
                    /* ... and search again in current block from beginning */
                    goto cont;
                }
            }
            bi->free_count -= (end - *beg);
            journal_mark_dirty (th, s, bi->bh);

            /* free block count calculation */
            reiserfs_prepare_for_journal (s, SB_BUFFER_WITH_SB(s), 1);
            PUT_SB_FREE_BLOCKS(s, SB_FREE_BLOCKS(s) - (end - *beg));
            journal_mark_dirty (th, s, SB_BUFFER_WITH_SB(s));

            return end - (*beg);
        } else {
            *beg = next;
        }
    }
}
Ejemplo n.º 27
0
/* Initializes an uninitialized block bitmap if given, and returns the
 * number of blocks free in the group. */
unsigned ext4_init_block_bitmap(struct super_block *sb, struct buffer_head *bh,
		 ext4_group_t block_group, struct ext4_group_desc *gdp)
{
	int bit, bit_max;
	ext4_group_t ngroups = ext4_get_groups_count(sb);
	unsigned free_blocks, group_blocks;
	struct ext4_sb_info *sbi = EXT4_SB(sb);

	if (bh) {
		J_ASSERT_BH(bh, buffer_locked(bh));

		/* If checksum is bad mark all blocks used to prevent allocation
		 * essentially implementing a per-group read-only flag. */
		if (!ext4_group_desc_csum_verify(sbi, block_group, gdp)) {
			ext4_error(sb, __func__,
				  "Checksum bad for group %u", block_group);
			ext4_free_blks_set(sb, gdp, 0);
			ext4_free_inodes_set(sb, gdp, 0);
			ext4_itable_unused_set(sb, gdp, 0);
			memset(bh->b_data, 0xff, sb->s_blocksize);
			return 0;
		}
		memset(bh->b_data, 0, sb->s_blocksize);
	}

	/* Check for superblock and gdt backups in this group */
	bit_max = ext4_bg_has_super(sb, block_group);

	if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_META_BG) ||
	    block_group < le32_to_cpu(sbi->s_es->s_first_meta_bg) *
			  sbi->s_desc_per_block) {
		if (bit_max) {
			bit_max += ext4_bg_num_gdb(sb, block_group);
			bit_max +=
				le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks);
		}
	} else { /* For META_BG_BLOCK_GROUPS */
		bit_max += ext4_bg_num_gdb(sb, block_group);
	}

	if (block_group == ngroups - 1) {
		/*
		 * Even though mke2fs always initialize first and last group
		 * if some other tool enabled the EXT4_BG_BLOCK_UNINIT we need
		 * to make sure we calculate the right free blocks
		 */
		group_blocks = ext4_blocks_count(sbi->s_es) -
			le32_to_cpu(sbi->s_es->s_first_data_block) -
			(EXT4_BLOCKS_PER_GROUP(sb) * (ngroups - 1));
	} else {
		group_blocks = EXT4_BLOCKS_PER_GROUP(sb);
	}

	free_blocks = group_blocks - bit_max;

	if (bh) {
		ext4_fsblk_t start, tmp;
		int flex_bg = 0;

		for (bit = 0; bit < bit_max; bit++)
			ext4_set_bit(bit, bh->b_data);

		start = ext4_group_first_block_no(sb, block_group);

		if (EXT4_HAS_INCOMPAT_FEATURE(sb,
					      EXT4_FEATURE_INCOMPAT_FLEX_BG))
			flex_bg = 1;

		/* Set bits for block and inode bitmaps, and inode table */
		tmp = ext4_block_bitmap(sb, gdp);
		if (!flex_bg || ext4_block_in_group(sb, tmp, block_group))
			ext4_set_bit(tmp - start, bh->b_data);

		tmp = ext4_inode_bitmap(sb, gdp);
		if (!flex_bg || ext4_block_in_group(sb, tmp, block_group))
			ext4_set_bit(tmp - start, bh->b_data);

		tmp = ext4_inode_table(sb, gdp);
		for (; tmp < ext4_inode_table(sb, gdp) +
				sbi->s_itb_per_group; tmp++) {
			if (!flex_bg ||
				ext4_block_in_group(sb, tmp, block_group))
				ext4_set_bit(tmp - start, bh->b_data);
		}
		/*
		 * Also if the number of blocks within the group is
		 * less than the blocksize * 8 ( which is the size
		 * of bitmap ), set rest of the block bitmap to 1
		 */
		mark_bitmap_end(group_blocks, sb->s_blocksize * 8, bh->b_data);
	}
	return free_blocks - ext4_group_used_meta_blocks(sb, block_group, gdp);
}
Ejemplo n.º 28
0
/*
 * 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 journal_commit_transaction(journal_t *journal)
{
	transaction_t *commit_transaction;
	struct journal_head *jh, *new_jh, *descriptor;
	struct buffer_head **wbuf = journal->j_wbuf;
	int bufs;
	int flags;
	int err;
	unsigned int blocknr;
	ktime_t start_time;
	u64 commit_time;
	char *tagp = NULL;
	journal_header_t *header;
	journal_block_tag_t *tag = NULL;
	int space_left = 0;
	int first_tag = 0;
	int tag_flag;
	int i;
	struct blk_plug plug;

	/*
	 * 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 journal_flush? */
	if (journal->j_flags & JFS_FLUSHED) {
		jbd_debug(3, "super block updated\n");
		journal_update_superblock(journal, 1);
	} 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;
	J_ASSERT(commit_transaction->t_state == T_RUNNING);

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

	spin_lock(&journal->j_state_lock);
	commit_transaction->t_state = T_LOCKED;

	trace_jbd_commit_locking(journal, commit_transaction);
	spin_lock(&commit_transaction->t_handle_lock);
	while (commit_transaction->t_updates) {
		DEFINE_WAIT(wait);

		prepare_to_wait(&journal->j_wait_updates, &wait,
					TASK_UNINTERRUPTIBLE);
		if (commit_transaction->t_updates) {
			spin_unlock(&commit_transaction->t_handle_lock);
			spin_unlock(&journal->j_state_lock);
			schedule();
			spin_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 (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 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 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 journal_get_undo_access()+journal_release_buffer() may
		 * leave undo-committed data.
		 */
		if (jh->b_committed_data) {
			struct buffer_head *bh = jh2bh(jh);

			jbd_lock_bh_state(bh);
			jbd_free(jh->b_committed_data, bh->b_size);
			jh->b_committed_data = NULL;
			jbd_unlock_bh_state(bh);
		}
		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);
	__journal_clean_checkpoint_list(journal);
	spin_unlock(&journal->j_list_lock);

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

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

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

	trace_jbd_commit_flushing(journal, commit_transaction);
	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);
	spin_unlock(&journal->j_state_lock);

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

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

	/*
	 * Wait for all previously submitted IO to complete.
	 */
	spin_lock(&journal->j_list_lock);
	while (commit_transaction->t_locked_list) {
		struct buffer_head *bh;

		jh = commit_transaction->t_locked_list->b_tprev;
		bh = jh2bh(jh);
		get_bh(bh);
		if (buffer_locked(bh)) {
			spin_unlock(&journal->j_list_lock);
			wait_on_buffer(bh);
			spin_lock(&journal->j_list_lock);
		}
		if (unlikely(!buffer_uptodate(bh))) {
			if (!trylock_page(bh->b_page)) {
				spin_unlock(&journal->j_list_lock);
				lock_page(bh->b_page);
				spin_lock(&journal->j_list_lock);
			}
			if (bh->b_page->mapping)
				set_bit(AS_EIO, &bh->b_page->mapping->flags);

			unlock_page(bh->b_page);
			SetPageError(bh->b_page);
			err = -EIO;
		}
		if (!inverted_lock(journal, bh)) {
			put_bh(bh);
			spin_lock(&journal->j_list_lock);
			continue;
		}
		if (buffer_jbd(bh) && bh2jh(bh) == jh &&
		    jh->b_transaction == commit_transaction &&
		    jh->b_jlist == BJ_Locked)
			__journal_unfile_buffer(jh);
		jbd_unlock_bh_state(bh);
		release_data_buffer(bh);
		cond_resched_lock(&journal->j_list_lock);
	}
	spin_unlock(&journal->j_list_lock);

	if (err) {
		char b[BDEVNAME_SIZE];

		printk(KERN_WARNING
			"JBD: Detected IO errors while flushing file data "
			"on %s\n", bdevname(journal->j_fs_dev, b));
		if (journal->j_flags & JFS_ABORT_ON_SYNCDATA_ERR)
			journal_abort(journal, err);
		err = 0;
	}

	blk_start_plug(&plug);

	journal_write_revoke_records(journal, commit_transaction, WRITE_SYNC);

	/*
	 * If we found any dirty or locked buffers, then we should have
	 * looped back up to the write_out_data label.  If there weren't
	 * any then journal_clean_data_list should have wiped the list
	 * clean by now, so check that it is in fact empty.
	 */
	J_ASSERT (commit_transaction->t_sync_datalist == NULL);

	jbd_debug (3, "JBD: commit phase 3\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:
	 */
	spin_lock(&journal->j_state_lock);
	commit_transaction->t_state = T_COMMIT;
	spin_unlock(&journal->j_state_lock);

	trace_jbd_commit_logging(journal, commit_transaction);
	J_ASSERT(commit_transaction->t_nr_buffers <=
		 commit_transaction->t_outstanding_credits);

	descriptor = NULL;
	bufs = 0;
	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");
			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) {
			struct buffer_head *bh;

			J_ASSERT (bufs == 0);

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

			descriptor = journal_get_descriptor_buffer(journal);
			if (!descriptor) {
				journal_abort(journal, -EIO);
				continue;
			}

			bh = jh2bh(descriptor);
			jbd_debug(4, "JBD: got buffer %llu (%p)\n",
				(unsigned long long)bh->b_blocknr, bh->b_data);
			header = (journal_header_t *)&bh->b_data[0];
			header->h_magic     = cpu_to_be32(JFS_MAGIC_NUMBER);
			header->h_blocktype = cpu_to_be32(JFS_DESCRIPTOR_BLOCK);
			header->h_sequence  = cpu_to_be32(commit_transaction->t_tid);

			tagp = &bh->b_data[sizeof(journal_header_t)];
			space_left = bh->b_size - sizeof(journal_header_t);
			first_tag = 1;
			set_buffer_jwrite(bh);
			set_buffer_dirty(bh);
			wbuf[bufs++] = bh;

			/* Record it so that we can wait for IO
                           completion later */
			BUFFER_TRACE(bh, "ph3: file as descriptor");
			journal_file_buffer(descriptor, commit_transaction,
					BJ_LogCtl);
		}

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

		err = 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) {
			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 journal_next_log_block() also.
		 */
		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 BJ_IO/BJ_Shadow pairing of buffers. */
		get_bh(jh2bh(jh));

		/* Make a temporary IO buffer with which to write it out
                   (this will requeue both the metadata buffer and the
                   temporary IO buffer). new_bh goes on BJ_IO*/

		set_buffer_jwrite(jh2bh(jh));
		/*
		 * akpm: journal_write_metadata_buffer() sets
		 * new_bh->b_transaction to commit_transaction.
		 * We need to clean this up before we release new_bh
		 * (which is of type BJ_IO)
		 */
		JBUFFER_TRACE(jh, "ph3: write metadata");
		flags = journal_write_metadata_buffer(commit_transaction,
						      jh, &new_jh, blocknr);
		set_buffer_jwrite(jh2bh(new_jh));
		wbuf[bufs++] = jh2bh(new_jh);

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

		tag_flag = 0;
		if (flags & 1)
			tag_flag |= JFS_FLAG_ESCAPE;
		if (!first_tag)
			tag_flag |= JFS_FLAG_SAME_UUID;

		tag = (journal_block_tag_t *) tagp;
		tag->t_blocknr = cpu_to_be32(jh2bh(jh)->b_blocknr);
		tag->t_flags = cpu_to_be32(tag_flag);
		tagp += sizeof(journal_block_tag_t);
		space_left -= sizeof(journal_block_tag_t);

		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 < sizeof(journal_block_tag_t) + 16) {

			jbd_debug(4, "JBD: 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_be32(JFS_FLAG_LAST_TAG);

start_journal_io:
			for (i = 0; i < bufs; i++) {
				struct buffer_head *bh = wbuf[i];
				lock_buffer(bh);
				clear_buffer_dirty(bh);
				set_buffer_uptodate(bh);
				bh->b_end_io = journal_end_buffer_io_sync;
				submit_bh(WRITE_SYNC, bh);
			}
			cond_resched();

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

	blk_finish_plug(&plug);

	/* Lo and behold: we have just managed to send a transaction to
           the log.  Before we can commit it, wait for the IO so far to
           complete.  Control buffers being written are on the
           transaction's t_log_list queue, and metadata buffers are on
           the t_iobuf_list queue.

	   Wait for the buffers in reverse order.  That way we are
	   less likely to be woken up until all IOs have completed, and
	   so we incur less scheduling load.
	*/

	jbd_debug(3, "JBD: commit phase 4\n");

	/*
	 * akpm: these are BJ_IO, and j_list_lock is not needed.
	 * See __journal_try_to_free_buffer.
	 */
wait_for_iobuf:
	while (commit_transaction->t_iobuf_list != NULL) {
		struct buffer_head *bh;

		jh = commit_transaction->t_iobuf_list->b_tprev;
		bh = jh2bh(jh);
		if (buffer_locked(bh)) {
			wait_on_buffer(bh);
			goto wait_for_iobuf;
		}
		if (cond_resched())
			goto wait_for_iobuf;

		if (unlikely(!buffer_uptodate(bh)))
			err = -EIO;

		clear_buffer_jwrite(bh);

		JBUFFER_TRACE(jh, "ph4: unfile after journal write");
		journal_unfile_buffer(journal, jh);

		/*
		 * ->t_iobuf_list should contain only dummy buffer_heads
		 * which were created by journal_write_metadata_buffer().
		 */
		BUFFER_TRACE(bh, "dumping temporary bh");
		journal_put_journal_head(jh);
		__brelse(bh);
		J_ASSERT_BH(bh, atomic_read(&bh->b_count) == 0);
		free_buffer_head(bh);

		/* We also have to unlock and free the corresponding
                   shadowed buffer */
		jh = commit_transaction->t_shadow_list->b_tprev;
		bh = jh2bh(jh);
		clear_buffer_jwrite(bh);
		J_ASSERT_BH(bh, buffer_jbddirty(bh));

		/* The metadata is now released for reuse, but we need
                   to remember it against this transaction so that when
                   we finally commit, we can do any checkpointing
                   required. */
		JBUFFER_TRACE(jh, "file as BJ_Forget");
		journal_file_buffer(jh, commit_transaction, BJ_Forget);
		/*
		 * Wake up any transactions which were waiting for this
		 * IO to complete. The barrier must be here so that changes
		 * by journal_file_buffer() take effect before wake_up_bit()
		 * does the waitqueue check.
		 */
		smp_mb();
		wake_up_bit(&bh->b_state, BH_Unshadow);
		JBUFFER_TRACE(jh, "brelse shadowed buffer");
		__brelse(bh);
	}

	J_ASSERT (commit_transaction->t_shadow_list == NULL);

	jbd_debug(3, "JBD: commit phase 5\n");

	/* Here we wait for the revoke record and descriptor record buffers */
 wait_for_ctlbuf:
	while (commit_transaction->t_log_list != NULL) {
		struct buffer_head *bh;

		jh = commit_transaction->t_log_list->b_tprev;
		bh = jh2bh(jh);
		if (buffer_locked(bh)) {
			wait_on_buffer(bh);
			goto wait_for_ctlbuf;
		}
		if (cond_resched())
			goto wait_for_ctlbuf;

		if (unlikely(!buffer_uptodate(bh)))
			err = -EIO;

		BUFFER_TRACE(bh, "ph5: control buffer writeout done: unfile");
		clear_buffer_jwrite(bh);
		journal_unfile_buffer(journal, jh);
		journal_put_journal_head(jh);
		__brelse(bh);		/* One for getblk */
		/* AKPM: bforget here */
	}

	if (err)
		journal_abort(journal, err);

	jbd_debug(3, "JBD: commit phase 6\n");

	/* All metadata is written, now write commit record and do cleanup */
	spin_lock(&journal->j_state_lock);
	J_ASSERT(commit_transaction->t_state == T_COMMIT);
	commit_transaction->t_state = T_COMMIT_RECORD;
	spin_unlock(&journal->j_state_lock);

	if (journal_write_commit_record(journal, commit_transaction))
		err = -EIO;

	if (err)
		journal_abort(journal, err);

	/* End of a transaction!  Finally, we can do checkpoint
           processing: any buffers committed as a result of this
           transaction can be removed from any checkpoint list it was on
           before. */

	jbd_debug(3, "JBD: commit phase 7\n");

	J_ASSERT(commit_transaction->t_sync_datalist == NULL);
	J_ASSERT(commit_transaction->t_buffers == NULL);
	J_ASSERT(commit_transaction->t_checkpoint_list == NULL);
	J_ASSERT(commit_transaction->t_iobuf_list == NULL);
	J_ASSERT(commit_transaction->t_shadow_list == NULL);
	J_ASSERT(commit_transaction->t_log_list == NULL);

restart_loop:
	/*
	 * As there are other places (journal_unmap_buffer()) adding buffers
	 * to this list we have to be careful and hold the j_list_lock.
	 */
	spin_lock(&journal->j_list_lock);
	while (commit_transaction->t_forget) {
		transaction_t *cp_transaction;
		struct buffer_head *bh;
		int try_to_free = 0;

		jh = commit_transaction->t_forget;
		spin_unlock(&journal->j_list_lock);
		bh = jh2bh(jh);
		/*
		 * Get a reference so that bh cannot be freed before we are
		 * done with it.
		 */
		get_bh(bh);
		jbd_lock_bh_state(bh);
		J_ASSERT_JH(jh,	jh->b_transaction == commit_transaction ||
			jh->b_transaction == journal->j_running_transaction);

		/*
		 * If there is undo-protected committed data against
		 * this buffer, then we can remove it now.  If it is a
		 * buffer needing such protection, the old frozen_data
		 * field now points to a committed version of the
		 * buffer, so rotate that field to the new committed
		 * data.
		 *
		 * Otherwise, we can just throw away the frozen data now.
		 */
		if (jh->b_committed_data) {
			jbd_free(jh->b_committed_data, bh->b_size);
			jh->b_committed_data = NULL;
			if (jh->b_frozen_data) {
				jh->b_committed_data = jh->b_frozen_data;
				jh->b_frozen_data = NULL;
			}
		} else if (jh->b_frozen_data) {
			jbd_free(jh->b_frozen_data, bh->b_size);
			jh->b_frozen_data = NULL;
		}

		spin_lock(&journal->j_list_lock);
		cp_transaction = jh->b_cp_transaction;
		if (cp_transaction) {
			JBUFFER_TRACE(jh, "remove from old cp transaction");
			__journal_remove_checkpoint(jh);
		}

		/* Only re-checkpoint the buffer_head if it is marked
		 * dirty.  If the buffer was added to the BJ_Forget list
		 * by journal_forget, it may no longer be dirty and
		 * there's no point in keeping a checkpoint record for
		 * it. */

		/* A buffer which has been freed while still being
		 * journaled by a previous transaction may end up still
		 * being dirty here, but we want to avoid writing back
		 * that buffer in the future after the "add to orphan"
		 * operation been committed,  That's not only a performance
		 * gain, it also stops aliasing problems if the buffer is
		 * left behind for writeback and gets reallocated for another
		 * use in a different page. */
		if (buffer_freed(bh) && !jh->b_next_transaction) {
			clear_buffer_freed(bh);
			clear_buffer_jbddirty(bh);
		}

		if (buffer_jbddirty(bh)) {
			JBUFFER_TRACE(jh, "add to new checkpointing trans");
			__journal_insert_checkpoint(jh, commit_transaction);
			if (is_journal_aborted(journal))
				clear_buffer_jbddirty(bh);
		} else {
			J_ASSERT_BH(bh, !buffer_dirty(bh));
			/*
			 * The buffer on BJ_Forget list and not jbddirty means
			 * it has been freed by this transaction and hence it
			 * could not have been reallocated until this
			 * transaction has committed. *BUT* it could be
			 * reallocated once we have written all the data to
			 * disk and before we process the buffer on BJ_Forget
			 * list.
			 */
			if (!jh->b_next_transaction)
				try_to_free = 1;
		}
		JBUFFER_TRACE(jh, "refile or unfile freed buffer");
		__journal_refile_buffer(jh);
		jbd_unlock_bh_state(bh);
		if (try_to_free)
			release_buffer_page(bh);
		else
			__brelse(bh);
		cond_resched_lock(&journal->j_list_lock);
	}
	spin_unlock(&journal->j_list_lock);
	/*
	 * This is a bit sleazy.  We use j_list_lock to protect transition
	 * of a transaction into T_FINISHED state and calling
	 * __journal_drop_transaction(). Otherwise we could race with
	 * other checkpointing code processing the transaction...
	 */
	spin_lock(&journal->j_state_lock);
	spin_lock(&journal->j_list_lock);
	/*
	 * Now recheck if some buffers did not get attached to the transaction
	 * while the lock was dropped...
	 */
	if (commit_transaction->t_forget) {
		spin_unlock(&journal->j_list_lock);
		spin_unlock(&journal->j_state_lock);
		goto restart_loop;
	}

	/* Done with this transaction! */

	jbd_debug(3, "JBD: commit phase 8\n");

	J_ASSERT(commit_transaction->t_state == T_COMMIT_RECORD);

	commit_transaction->t_state = T_FINISHED;
	J_ASSERT(commit_transaction == journal->j_committing_transaction);
	journal->j_commit_sequence = commit_transaction->t_tid;
	journal->j_committing_transaction = NULL;
	commit_time = ktime_to_ns(ktime_sub(ktime_get(), start_time));

	/*
	 * weight the commit time higher than the average time so we don't
	 * react too strongly to vast changes in commit time
	 */
	if (likely(journal->j_average_commit_time))
		journal->j_average_commit_time = (commit_time*3 +
				journal->j_average_commit_time) / 4;
	else
		journal->j_average_commit_time = commit_time;

	spin_unlock(&journal->j_state_lock);

	if (commit_transaction->t_checkpoint_list == NULL &&
	    commit_transaction->t_checkpoint_io_list == NULL) {
		__journal_drop_transaction(journal, commit_transaction);
	} else {
		if (journal->j_checkpoint_transactions == NULL) {
			journal->j_checkpoint_transactions = commit_transaction;
			commit_transaction->t_cpnext = commit_transaction;
			commit_transaction->t_cpprev = commit_transaction;
		} else {
			commit_transaction->t_cpnext =
				journal->j_checkpoint_transactions;
			commit_transaction->t_cpprev =
				commit_transaction->t_cpnext->t_cpprev;
			commit_transaction->t_cpnext->t_cpprev =
				commit_transaction;
			commit_transaction->t_cpprev->t_cpnext =
				commit_transaction;
		}
	}
	spin_unlock(&journal->j_list_lock);

	trace_jbd_end_commit(journal, commit_transaction);
	jbd_debug(1, "JBD: commit %d complete, head %d\n",
		  journal->j_commit_sequence, journal->j_tail_sequence);

	wake_up(&journal->j_wait_done_commit);
}
Ejemplo n.º 29
0
/*
 *  Submit all the data buffers to disk
 */
static int journal_submit_data_buffers(journal_t *journal,
				       transaction_t *commit_transaction,
				       int write_op)
{
	struct journal_head *jh;
	struct buffer_head *bh;
	int locked;
	int bufs = 0;
	struct buffer_head **wbuf = journal->j_wbuf;
	int err = 0;

	/*
	 * Whenever we unlock the journal and sleep, things can get added
	 * onto ->t_sync_datalist, so we have to keep looping back to
	 * write_out_data until we *know* that the list is empty.
	 *
	 * Cleanup any flushed data buffers from the data list.  Even in
	 * abort mode, we want to flush this out as soon as possible.
	 */
write_out_data:
	cond_resched();
	spin_lock(&journal->j_list_lock);

	while (commit_transaction->t_sync_datalist) {
		jh = commit_transaction->t_sync_datalist;
		bh = jh2bh(jh);
		locked = 0;

		/* Get reference just to make sure buffer does not disappear
		 * when we are forced to drop various locks */
		get_bh(bh);
		/* If the buffer is dirty, we need to submit IO and hence
		 * we need the buffer lock. We try to lock the buffer without
		 * blocking. If we fail, we need to drop j_list_lock and do
		 * blocking lock_buffer().
		 */
		if (buffer_dirty(bh)) {
			if (!trylock_buffer(bh)) {
				BUFFER_TRACE(bh, "needs blocking lock");
				spin_unlock(&journal->j_list_lock);
				trace_jbd_do_submit_data(journal,
						     commit_transaction);
				/* Write out all data to prevent deadlocks */
				journal_do_submit_data(wbuf, bufs, write_op);
				bufs = 0;
				lock_buffer(bh);
				spin_lock(&journal->j_list_lock);
			}
			locked = 1;
		}
		/* We have to get bh_state lock. Again out of order, sigh. */
		if (!inverted_lock(journal, bh)) {
			jbd_lock_bh_state(bh);
			spin_lock(&journal->j_list_lock);
		}
		/* Someone already cleaned up the buffer? */
		if (!buffer_jbd(bh) || bh2jh(bh) != jh
			|| jh->b_transaction != commit_transaction
			|| jh->b_jlist != BJ_SyncData) {
			jbd_unlock_bh_state(bh);
			if (locked)
				unlock_buffer(bh);
			BUFFER_TRACE(bh, "already cleaned up");
			release_data_buffer(bh);
			continue;
		}
		if (locked && test_clear_buffer_dirty(bh)) {
			BUFFER_TRACE(bh, "needs writeout, adding to array");
			wbuf[bufs++] = bh;
			__journal_file_buffer(jh, commit_transaction,
						BJ_Locked);
			jbd_unlock_bh_state(bh);
			if (bufs == journal->j_wbufsize) {
				spin_unlock(&journal->j_list_lock);
				trace_jbd_do_submit_data(journal,
						     commit_transaction);
				journal_do_submit_data(wbuf, bufs, write_op);
				bufs = 0;
				goto write_out_data;
			}
		} else if (!locked && buffer_locked(bh)) {
			__journal_file_buffer(jh, commit_transaction,
						BJ_Locked);
			jbd_unlock_bh_state(bh);
			put_bh(bh);
		} else {
			BUFFER_TRACE(bh, "writeout complete: unfile");
			if (unlikely(!buffer_uptodate(bh)))
				err = -EIO;
			__journal_unfile_buffer(jh);
			jbd_unlock_bh_state(bh);
			if (locked)
				unlock_buffer(bh);
			release_data_buffer(bh);
		}

		if (need_resched() || spin_needbreak(&journal->j_list_lock)) {
			spin_unlock(&journal->j_list_lock);
			goto write_out_data;
		}
	}
	spin_unlock(&journal->j_list_lock);
	trace_jbd_do_submit_data(journal, commit_transaction);
	journal_do_submit_data(wbuf, bufs, write_op);

	return err;
}
Ejemplo n.º 30
0
int journal_get_create_access(handle_t *handle, struct buffer_head *bh)
{
	transaction_t *transaction = handle->h_transaction;
	journal_t *journal = transaction->t_journal;
	struct journal_head *jh = journal_add_journal_head(bh);
	int err;

	jbd_debug(5, "journal_head %p\n", jh);
	err = -EROFS;
	if (is_handle_aborted(handle))
		goto out;
	err = 0;

	JBUFFER_TRACE(jh, "entry");
	/*
	 * The buffer may already belong to this transaction due to pre-zeroing
	 * in the filesystem's new_block code.  It may also be on the previous,
	 * committing transaction's lists, but it HAS to be in Forget state in
	 * that case: the transaction must have deleted the buffer for it to be
	 * reused here.
	 */
	jbd_lock_bh_state(bh);
	spin_lock(&journal->j_list_lock);
	J_ASSERT_JH(jh, (jh->b_transaction == transaction ||
		jh->b_transaction == NULL ||
		(jh->b_transaction == journal->j_committing_transaction &&
			  jh->b_jlist == BJ_Forget)));

	J_ASSERT_JH(jh, jh->b_next_transaction == NULL);
	J_ASSERT_JH(jh, buffer_locked(jh2bh(jh)));

	if (jh->b_transaction == NULL) {
		/*
		 * Previous journal_forget() could have left the buffer
		 * with jbddirty bit set because it was being committed. When
		 * the commit finished, we've filed the buffer for
		 * checkpointing and marked it dirty. Now we are reallocating
		 * the buffer so the transaction freeing it must have
		 * committed and so it's safe to clear the dirty bit.
		 */
		clear_buffer_dirty(jh2bh(jh));
		jh->b_transaction = transaction;

		/* first access by this transaction */
		jh->b_modified = 0;

		JBUFFER_TRACE(jh, "file as BJ_Reserved");
		__journal_file_buffer(jh, transaction, BJ_Reserved);
	} else if (jh->b_transaction == journal->j_committing_transaction) {
		/* first access by this transaction */
		jh->b_modified = 0;

		JBUFFER_TRACE(jh, "set next transaction");
		jh->b_next_transaction = transaction;
	}
	spin_unlock(&journal->j_list_lock);
	jbd_unlock_bh_state(bh);

	/*
	 * akpm: I added this.  ext3_alloc_branch can pick up new indirect
	 * blocks which contain freed but then revoked metadata.  We need
	 * to cancel the revoke in case we end up freeing it yet again
	 * and the reallocating as data - this would cause a second revoke,
	 * which hits an assertion error.
	 */
	JBUFFER_TRACE(jh, "cancelling revoke");
	journal_cancel_revoke(handle, jh);
	journal_put_journal_head(jh);
out:
	return err;
}