replay.c

可以在不启动LINUX的情况下直接访问EXT2和EXT3格式的磁盘

#include <linux/module.h>
#include <linux/time.h>
#include <linux/fs.h>
#include <linux/jbd.h>
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/mm.h>
#include <linux/freezer.h>
#include <linux/pagemap.h>
#include <linux/kthread.h>
#include <linux/poison.h>
#include <linux/proc_fs.h>
#include <linux/debugfs.h>


/*
 * Called under j_state_lock.  Returns true if a transaction was started.
 */
int __log_start_commit(journal_t *journal, tid_t target)
{
	/*
	 * Are we already doing a recent enough commit?
	 */
	if (!tid_geq(journal->j_commit_request, target)) {
		/*
		 * We want a new commit: OK, mark the request and wakup the
		 * commit thread.  We do _not_ do the commit ourselves.
		 */

		journal->j_commit_request = target;
		jbd_debug(1, "JBD: requesting commit %d/%d\n",
			  journal->j_commit_request,
			  journal->j_commit_sequence);
		wake_up(&journal->j_wait_commit);
		return 1;
	}
	return 0;
}

int log_start_commit(journal_t *journal, tid_t tid)
{
	int ret;

	spin_lock(&journal->j_state_lock);
	ret = __log_start_commit(journal, tid);
	spin_unlock(&journal->j_state_lock);
	return ret;
}

/*
 * Journal abort has very specific semantics, which we describe
 * for journal abort.
 *
 * Two internal function, which provide abort to te jbd layer
 * 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.
 */
static void __journal_abort_hard(journal_t *journal)
{
	transaction_t *transaction;

	if (journal->j_flags & JFS_ABORT)
		return;

	spin_lock(&journal->j_state_lock);
	journal->j_flags |= JFS_ABORT;
	transaction = journal->j_running_transaction;
	if (transaction)
		__log_start_commit(journal, transaction->t_tid);
	spin_unlock(&journal->j_state_lock);
}

/* 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 err)
{
	if (journal->j_flags & JFS_ABORT)
		return;

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

	__journal_abort_hard(journal);

	if (err)
		journal_update_superblock(journal, 1);
}


/**
 * void journal_abort () - Shutdown the journal immediately.
 * @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.
 *
 * The journal_abort function is intended to support higher level error
 * 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
 * journal_stop on an existing handle will return -EIO if we have
 * entered abort state during the update.
 *
 * Recursive transactions are not disturbed by journal abort until the
 * final journal_stop, which will receive the -EIO error.
 *
 * Finally, the journal_abort call allows the caller to supply an errno
 * 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).
 *
 */

void journal_abort(journal_t *journal, int err)
{
	__journal_abort_soft(journal, err);
}

/**
 * int journal_errno () - returns the journal's error state.
 * @journal: journal to examine.
 *
 * This is the errno numbet set with journal_abort(), the last
 * 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.
 */
int journal_errno(journal_t *journal)
{
	int err;

	spin_lock(&journal->j_state_lock);
	if (journal->j_flags & JFS_ABORT)
		err = -EROFS;
	else
		err = journal->j_errno;
	spin_unlock(&journal->j_state_lock);
	return err;
}

/**
 * int journal_clear_err () - clears the journal's error state
 * @journal: journal to act on.
 *
 * An error must be cleared or Acked to take a FS out of readonly
 * mode.
 */
int journal_clear_err(journal_t *journal)
{
	int err = 0;

	spin_lock(&journal->j_state_lock);
	if (journal->j_flags & JFS_ABORT)
		err = -EROFS;
	else
		journal->j_errno = 0;
	spin_unlock(&journal->j_state_lock);
	return err;
}

/**
 * void journal_ack_err() - Ack journal err.
 * @journal: journal to act on.
 *
 * An error must be cleared or Acked to take a FS out of readonly
 * mode.
 */
void journal_ack_err(journal_t *journal)
{
	spin_lock(&journal->j_state_lock);
	if (journal->j_errno)
		journal->j_flags |= JFS_ACK_ERR;
	spin_unlock(&journal->j_state_lock);
}

int journal_blocks_per_page(struct inode *inode)
{
	return 1 << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits);
}


/*
 * Journal_head storage management
 */
static struct kmem_cache *journal_head_cache = NULL;
#ifdef CONFIG_JBD_DEBUG
static atomic_t nr_journal_heads = ATOMIC_INIT(0);
#endif

static int journal_init_journal_head_cache(void)
{
	int retval;

	J_ASSERT(journal_head_cache == 0);
	journal_head_cache = kmem_cache_create("journal_head",
				sizeof(struct journal_head),
				0,		/* offset */
				SLAB_TEMPORARY,	/* flags */
				NULL);		/* ctor */
	retval = 0;
	if (journal_head_cache == 0) {
		retval = -ENOMEM;
		printk(KERN_EMERG "JBD: no memory for journal_head cache\n");
	}
	return retval;
}

static void journal_destroy_journal_head_cache(void)
{
	J_ASSERT(journal_head_cache != NULL);
	kmem_cache_destroy(journal_head_cache);
	journal_head_cache = NULL;
}

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

#ifdef CONFIG_JBD_DEBUG
	atomic_inc(&nr_journal_heads);
#endif
	ret = kmem_cache_alloc(journal_head_cache, GFP_NOFS);
	if (ret == NULL) {
		jbd_debug(1, "out of memory for journal_head\n");
		if (time_after(jiffies, last_warning + 5*HZ)) {
			printk(KERN_NOTICE "ENOMEM in %s, retrying.\n",
			       __FUNCTION__);
			last_warning = jiffies;
		}
		while (ret == NULL) {
			yield();
			ret = kmem_cache_alloc(journal_head_cache, GFP_NOFS);
		}
	}
	return ret;
}

static void journal_free_journal_head(struct journal_head *jh)
{
#ifdef CONFIG_JBD_DEBUG
	atomic_dec(&nr_journal_heads);
	memset(jh, JBD_POISON_FREE, sizeof(*jh));
#endif
	kmem_cache_free(journal_head_cache, jh);
}

/*
 * 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.
 *
 * A journal_head may be detached from its buffer_head when the journal_head's
 * b_transaction, b_cp_transaction and b_next_transaction pointers are NULL.
 * Various places in JBD call journal_remove_journal_head() to indicate that the
 * journal_head can be dropped if needed.
 *
 * 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
 * journal_head in this situation, journal_add_journal_head elevates the
 * journal_head's b_jcount refcount by one.  The caller must call
 * journal_put_journal_head() to undo this.
 *
 * So the typical usage would be:
 *
 *	(Attach a journal_head if needed.  Increments b_jcount)
 *	struct journal_head *jh = journal_add_journal_head(bh);
 *	...
 *	jh->b_transaction = xxx;
 *	journal_put_journal_head(jh);
 *
 * Now, the journal_head's b_jcount is zero, but it is safe from being released
 * because it has a non-zero b_transaction.
 */

/*
 * Give a buffer_head a journal_head.
 *
 * Doesn't need the journal lock.
 * May sleep.
 */
struct journal_head *journal_add_journal_head(struct buffer_head *bh)
{
	struct journal_head *jh;
	struct journal_head *new_jh = NULL;

repeat:
	if (!buffer_jbd(bh)) {
		new_jh = journal_alloc_journal_head();
		memset(new_jh, 0, sizeof(*new_jh));
	}

	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
 */
struct journal_head *journal_grab_journal_head(struct buffer_head *bh)
{
	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);

	get_bh(bh);
	if (jh->b_jcount == 0) {
		if (jh->b_transaction == NULL &&
				jh->b_next_transaction == NULL &&
				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",
						__FUNCTION__);
				jbd_free(jh->b_frozen_data, bh->b_size);
			}
			if (jh->b_committed_data) {
				printk(KERN_WARNING "%s: freeing "
						"b_committed_data\n",
						__FUNCTION__);
				jbd_free(jh->b_committed_data, bh->b_size);
			}
			bh->b_private = NULL;
			jh->b_bh = NULL;	/* debug, really */
			clear_buffer_jbd(bh);
			__brelse(bh);
			journal_free_journal_head(jh);
		} else {
			BUFFER_TRACE(bh, "journal_head was locked");
		}
	}
}

/*
 * journal_remove_journal_head(): if the buffer isn't attached to a transaction
 * and has a zero b_jcount then remove and release its journal_head.   If we did
 * see that the buffer is not used by any transaction we also "logically"
 * decrement ->b_count.
 *
 * We in fact take an additional increment on ->b_count as a convenience,
 * because the caller usually wants to do additional things with the bh
 * after calling here.
 * The caller of journal_remove_journal_head() *must* run __brelse(bh) at some
 * time.  Once the caller has run __brelse(), the buffer is eligible for
 * reaping by try_to_free_buffers().
 */
void journal_remove_journal_head(struct buffer_head *bh)
{
	jbd_lock_bh_journal_head(bh);
	__journal_remove_journal_head(bh);
	jbd_unlock_bh_journal_head(bh);
}

/*
 * Drop a reference on the passed journal_head.  If it fell to zero then try to
 * release the journal_head from the buffer_head.
 */
void journal_put_journal_head(struct journal_head *jh)
{
	struct buffer_head *bh = jh2bh(jh);

	jbd_lock_bh_journal_head(bh);
	J_ASSERT_JH(jh, jh->b_jcount > 0);
	--jh->b_jcount;
	if (!jh->b_jcount && !jh->b_transaction) {
		__journal_remove_journal_head(bh);
		__brelse(bh);
	}
	jbd_unlock_bh_journal_head(bh);
}

/*
 * Log buffer allocation routines:
 */

int journal_next_log_block(journal_t *journal, unsigned long *retp)
{
	unsigned long blocknr;

	spin_lock(&journal->j_state_lock);
	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;
	spin_unlock(&journal->j_state_lock);
	return journal_bmap(journal, blocknr, retp);
}

/*
 * 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.
 */
int journal_bmap(journal_t *journal, unsigned long blocknr,
		 unsigned long *retp)
{
	int err = 0;
	unsigned long ret;

	if (journal->j_inode) {
		ret = (unsigned long)bmap(journal->j_inode, (sector_t)blocknr);
		if (ret)
			*retp = ret;
		else {
			printk(KERN_ALERT "%s: journal block not found "
					"at offset %lu ...\n",
				__FUNCTION__,
				blocknr);
			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.
 *
 * After the caller of journal_get_descriptor_buffer() has finished modifying
 * 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.
 */
struct journal_head *journal_get_descriptor_buffer(journal_t *journal)
{
	struct buffer_head *bh;
	unsigned long blocknr;
	int err;

	err = journal_next_log_block(journal, &blocknr);

	if (err)
		return NULL;

	bh = __getblk(journal->j_dev, blocknr, journal->j_blocksize);
	lock_buffer(bh);
	memset(bh->b_data, 0, journal->j_blocksize);
	set_buffer_uptodate(bh);
	unlock_buffer(bh);
	BUFFER_TRACE(bh, "return this buffer");
	return journal_add_journal_head(bh);
}

/*
 * 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)
{
	journal_t *journal;
	int err;

	journal = kzalloc(sizeof(*journal), GFP_KERNEL);