1 files changed, 164 insertions, 161 deletions

diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c
index 6914598022c..7f34f471616 100644
--- a/fs/jbd2/checkpoint.c
+++ b/fs/jbd2/checkpoint.c
@@ -22,6 +22,8 @@
 #include <linux/jbd2.h>
 #include <linux/errno.h>
 #include <linux/slab.h>
+#include <linux/blkdev.h>
+#include <trace/events/jbd2.h>
 
 /*
  * Unlink a buffer from a transaction checkpoint list.
@@ -86,22 +88,23 @@ static inline void __buffer_relink_io(struct journal_head *jh)
  * 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)) {
+	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;
-		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;
 }
@@ -114,79 +117,103 @@ static int __try_to_free_cp_buf(struct journal_head *jh)
  */
 void __jbd2_log_wait_for_space(journal_t *journal)
 {
-	int nblocks;
-	assert_spin_locked(&journal->j_state_lock);
+	int nblocks, space_left;
+	/* assert_spin_locked(&journal->j_state_lock); */
 
-	nblocks = jbd_space_needed(journal);
-	while (__jbd2_log_space_left(journal) < nblocks) {
+	nblocks = jbd2_space_needed(journal);
+	while (jbd2_log_space_left(journal) < nblocks) {
 		if (journal->j_flags & JBD2_ABORT)
 			return;
-		spin_unlock(&journal->j_state_lock);
+		write_unlock(&journal->j_state_lock);
 		mutex_lock(&journal->j_checkpoint_mutex);
 
 		/*
 		 * Test again, another process may have checkpointed while we
-		 * were waiting for the checkpoint lock
+		 * were waiting for the checkpoint lock. If there are no
+		 * transactions ready to be checkpointed, try to recover
+		 * journal space by calling cleanup_journal_tail(), and if
+		 * that doesn't work, by waiting for the currently committing
+		 * transaction to complete.  If there is absolutely no way
+		 * to make progress, this is either a BUG or corrupted
+		 * filesystem, so abort the journal and leave a stack
+		 * trace for forensic evidence.
 		 */
-		spin_lock(&journal->j_state_lock);
-		nblocks = jbd_space_needed(journal);
-		if (__jbd2_log_space_left(journal) < nblocks) {
-			spin_unlock(&journal->j_state_lock);
-			jbd2_log_do_checkpoint(journal);
-			spin_lock(&journal->j_state_lock);
+		write_lock(&journal->j_state_lock);
+		spin_lock(&journal->j_list_lock);
+		nblocks = jbd2_space_needed(journal);
+		space_left = jbd2_log_space_left(journal);
+		if (space_left < nblocks) {
+			int chkpt = journal->j_checkpoint_transactions != NULL;
+			tid_t tid = 0;
+
+			if (journal->j_committing_transaction)
+				tid = journal->j_committing_transaction->t_tid;
+			spin_unlock(&journal->j_list_lock);
+			write_unlock(&journal->j_state_lock);
+			if (chkpt) {
+				jbd2_log_do_checkpoint(journal);
+			} else if (jbd2_cleanup_journal_tail(journal) == 0) {
+				/* We were able to recover space; yay! */
+				;
+			} else if (tid) {
+				/*
+				 * jbd2_journal_commit_transaction() may want
+				 * to take the checkpoint_mutex if JBD2_FLUSHED
+				 * is set.  So we need to temporarily drop it.
+				 */
+				mutex_unlock(&journal->j_checkpoint_mutex);
+				jbd2_log_wait_commit(journal, tid);
+				write_lock(&journal->j_state_lock);
+				continue;
+			} else {
+				printk(KERN_ERR "%s: needed %d blocks and "
+				       "only had %d space available\n",
+				       __func__, nblocks, space_left);
+				printk(KERN_ERR "%s: no way to get more "
+				       "journal space in %s\n", __func__,
+				       journal->j_devname);
+				WARN_ON(1);
+				jbd2_journal_abort(journal, 0);
+			}
+			write_lock(&journal->j_state_lock);
+		} else {
+			spin_unlock(&journal->j_list_lock);
 		}
 		mutex_unlock(&journal->j_checkpoint_mutex);
 	}
 }
 
 /*
- * We were unable to perform jbd_trylock_bh_state() inside j_list_lock.
- * The caller must restart a list walk.  Wait for someone else to run
- * jbd_unlock_bh_state().
- */
-static void jbd_sync_bh(journal_t *journal, struct buffer_head *bh)
-	__releases(journal->j_list_lock)
-{
-	get_bh(bh);
-	spin_unlock(&journal->j_list_lock);
-	jbd_lock_bh_state(bh);
-	jbd_unlock_bh_state(bh);
-	put_bh(bh);
-}
-
-/*
  * 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 void __wait_cp_io(journal_t *journal, transaction_t *transaction)
+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;
+		return ret;
 	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;
-		}
+		get_bh(bh);
 		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");
@@ -194,28 +221,33 @@ restart:
 			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);
-		jbd_unlock_bh_state(bh);
-		jbd2_journal_remove_journal_head(bh);
 		__brelse(bh);
 	}
-}
 
-#define NR_BATCH	64
+	return ret;
+}
 
 static void
-__flush_batch(journal_t *journal, struct buffer_head **bhs, int *batch_count)
+__flush_batch(journal_t *journal, int *batch_count)
 {
 	int i;
+	struct blk_plug plug;
+
+	blk_start_plug(&plug);
+	for (i = 0; i < *batch_count; i++)
+		write_dirty_buffer(journal->j_chkpt_bhs[i], WRITE_SYNC);
+	blk_finish_plug(&plug);
 
-	ll_rw_block(SWRITE, *batch_count, bhs);
 	for (i = 0; i < *batch_count; i++) {
-		struct buffer_head *bh = bhs[i];
-		clear_buffer_jwrite(bh);
+		struct buffer_head *bh = journal->j_chkpt_bhs[i];
 		BUFFER_TRACE(bh, "brelse");
 		__brelse(bh);
 	}
@@ -226,22 +258,20 @@ __flush_batch(journal_t *journal, struct buffer_head **bhs, int *batch_count)
  * 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.
+ * scan of the checkpoint list.  Return <0 if the buffer has failed to
+ * be written out.
  *
  * 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,
-			transaction_t *transaction)
+			    int *batch_count, transaction_t *transaction)
 {
 	struct buffer_head *bh = jh2bh(jh);
 	int ret = 0;
 
 	if (buffer_locked(bh)) {
-		atomic_inc(&bh->b_count);
+		get_bh(bh);
 		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");
@@ -253,19 +283,28 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh,
 
 		transaction->t_chp_stats.cs_forced_to_close++;
 		spin_unlock(&journal->j_list_lock);
-		jbd_unlock_bh_state(bh);
+		if (unlikely(journal->j_flags & JBD2_UNMOUNT))
+			/*
+			 * The journal thread is dead; so starting and
+			 * waiting for a commit to finish will cause
+			 * us to wait for a _very_ long time.
+			 */
+			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)) {
-		J_ASSERT_JH(jh, !buffer_jbddirty(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);
-		jbd_unlock_bh_state(bh);
-		jbd2_journal_remove_journal_head(bh);
 		__brelse(bh);
-		ret = 1;
 	} else {
 		/*
 		 * Important: we are about to write the buffer, and
@@ -277,15 +316,13 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh,
 		BUFFER_TRACE(bh, "queue");
 		get_bh(bh);
 		J_ASSERT_BH(bh, !buffer_jwrite(bh));
-		set_buffer_jwrite(bh);
-		bhs[*batch_count] = bh;
+		journal->j_chkpt_bhs[*batch_count] = bh;
 		__buffer_relink_io(jh);
-		jbd_unlock_bh_state(bh);
 		transaction->t_chp_stats.cs_written++;
 		(*batch_count)++;
-		if (*batch_count == NR_BATCH) {
+		if (*batch_count == JBD2_NR_BATCH) {
 			spin_unlock(&journal->j_list_lock);
-			__flush_batch(journal, bhs, batch_count);
+			__flush_batch(journal, batch_count);
 			ret = 1;
 		}
 	}
@@ -298,6 +335,7 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh,
  * to disk. We submit larger chunks of data at once.
  *
  * The journal should be locked before calling this function.
+ * Called with j_checkpoint_mutex held.
  */
 int jbd2_log_do_checkpoint(journal_t *journal)
 {
@@ -313,6 +351,7 @@ int jbd2_log_do_checkpoint(journal_t *journal)
 	 * journal straight away.
 	 */
 	result = jbd2_cleanup_journal_tail(journal);
+	trace_jbd2_checkpoint(journal, result);
 	jbd_debug(1, "cleanup_journal_tail returned %d\n", result);
 	if (result <= 0)
 		return result;
@@ -321,6 +360,7 @@ int jbd2_log_do_checkpoint(journal_t *journal)
 	 * OK, we need to start writing disk blocks.  Take one transaction
 	 * and write it.
 	 */
+	result = 0;
 	spin_lock(&journal->j_list_lock);
 	if (!journal->j_checkpoint_transactions)
 		goto out;
@@ -337,22 +377,15 @@ restart:
 	if (journal->j_checkpoint_transactions == transaction &&
 			transaction->t_tid == this_tid) {
 		int batch_count = 0;
-		struct buffer_head *bhs[NR_BATCH];
 		struct journal_head *jh;
-		int retry = 0;
+		int retry = 0, err;
 
 		while (!retry && transaction->t_checkpoint_list) {
-			struct buffer_head *bh;
-
 			jh = transaction->t_checkpoint_list;
-			bh = jh2bh(jh);
-			if (!jbd_trylock_bh_state(bh)) {
-				jbd_sync_bh(journal, bh);
-				retry = 1;
-				break;
-			}
-			retry = __process_buffer(journal, jh, bhs, &batch_count,
+			retry = __process_buffer(journal, jh, &batch_count,
 						 transaction);
+			if (retry < 0 && !result)
+				result = retry;
 			if (!retry && (need_resched() ||
 				spin_needbreak(&journal->j_list_lock))) {
 				spin_unlock(&journal->j_list_lock);
@@ -366,7 +399,7 @@ restart:
 				spin_unlock(&journal->j_list_lock);
 				retry = 1;
 			}
-			__flush_batch(journal, bhs, &batch_count);
+			__flush_batch(journal, &batch_count);
 		}
 
 		if (retry) {
@@ -377,14 +410,18 @@ restart:
 		 * Now we have cleaned up the first transaction's checkpoint
 		 * list. Let's clean up the second one
 		 */
-		__wait_cp_io(journal, transaction);
+		err = __wait_cp_io(journal, transaction);
+		if (!result)
+			result = err;
 	}
 out:
 	spin_unlock(&journal->j_list_lock);
-	result = jbd2_cleanup_journal_tail(journal);
 	if (result < 0)
-		return result;
-	return 0;
+		jbd2_journal_abort(journal, result);
+	else
+		result = jbd2_cleanup_journal_tail(journal);
+
+	return (result < 0) ? result : 0;
 }
 
 /*
@@ -400,67 +437,35 @@ out:
  * This is the only part of the journaling code which really needs to be
  * aware of transaction aborts.  Checkpointing involves writing to the
  * main filesystem area rather than to the journal, so it can proceed
- * even in abort state, but we must not update the journal superblock if
- * we have an abort error outstanding.
+ * even in abort state, but we must not update the super block if
+ * checkpointing may have failed.  Otherwise, we would lose some metadata
+ * buffers which should be written-back to the filesystem.
  */
 
 int jbd2_cleanup_journal_tail(journal_t *journal)
 {
-	transaction_t * transaction;
 	tid_t		first_tid;
-	unsigned long	blocknr, freed;
+	unsigned long	blocknr;
 
-	/* OK, work out the oldest transaction remaining in the log, and
-	 * the log block it starts at.
-	 *
-	 * If the log is now empty, we need to work out which is the
-	 * next transaction ID we will write, and where it will
-	 * start. */
+	if (is_journal_aborted(journal))
+		return 1;
 
-	spin_lock(&journal->j_state_lock);
-	spin_lock(&journal->j_list_lock);
-	transaction = journal->j_checkpoint_transactions;
-	if (transaction) {
-		first_tid = transaction->t_tid;
-		blocknr = transaction->t_log_start;
-	} else if ((transaction = journal->j_committing_transaction) != NULL) {
-		first_tid = transaction->t_tid;
-		blocknr = transaction->t_log_start;
-	} else if ((transaction = journal->j_running_transaction) != NULL) {
-		first_tid = transaction->t_tid;
-		blocknr = journal->j_head;
-	} else {
-		first_tid = journal->j_transaction_sequence;
-		blocknr = journal->j_head;
-	}
-	spin_unlock(&journal->j_list_lock);
+	if (!jbd2_journal_get_log_tail(journal, &first_tid, &blocknr))
+		return 1;
 	J_ASSERT(blocknr != 0);
 
-	/* If the oldest pinned transaction is at the tail of the log
-           already then there's not much we can do right now. */
-	if (journal->j_tail_sequence == first_tid) {
-		spin_unlock(&journal->j_state_lock);
-		return 1;
-	}
+	/*
+	 * We need to make sure that any blocks that were recently written out
+	 * --- perhaps by jbd2_log_do_checkpoint() --- are flushed out before
+	 * we drop the transactions from the journal. It's unlikely this will
+	 * be necessary, especially with an appropriately sized journal, but we
+	 * need this to guarantee correctness.  Fortunately
+	 * jbd2_cleanup_journal_tail() doesn't get called all that often.
+	 */
+	if (journal->j_flags & JBD2_BARRIER)
+		blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL);
 
-	/* OK, update the superblock to recover the freed space.
-	 * Physical blocks come first: have we wrapped beyond the end of
-	 * the log?  */
-	freed = blocknr - journal->j_tail;
-	if (blocknr < journal->j_tail)
-		freed = freed + journal->j_last - journal->j_first;
-
-	jbd_debug(1,
-		  "Cleaning journal tail from %d to %d (offset %lu), "
-		  "freeing %lu\n",
-		  journal->j_tail_sequence, first_tid, blocknr, freed);
-
-	journal->j_free += freed;
-	journal->j_tail_sequence = first_tid;
-	journal->j_tail = blocknr;
-	spin_unlock(&journal->j_state_lock);
-	if (!(journal->j_flags & JBD2_ABORT))
-		jbd2_journal_update_superblock(journal, 1);
+	__jbd2_update_log_tail(journal, first_tid, blocknr);
 	return 0;
 }
 
@@ -470,11 +475,12 @@ int jbd2_cleanup_journal_tail(journal_t *journal)
 /*
  * journal_clean_one_cp_list
  *
- * Find all the written-back checkpoint buffers in the given list and release them.
+ * Find all the written-back checkpoint buffers in the given list and
+ * release them.
  *
  * Called with the journal locked.
  * Called with j_list_lock held.
- * Returns number of bufers reaped (for debug)
+ * Returns number of buffers reaped (for debug)
  */
 
 static int journal_clean_one_cp_list(struct journal_head *jh, int *released)
@@ -491,15 +497,12 @@ static int journal_clean_one_cp_list(struct journal_head *jh, int *released)
 	do {
 		jh = next_jh;
 		next_jh = jh->b_cpnext;
-		/* Use trylock because of the ranking */
-		if (jbd_trylock_bh_state(jh2bh(jh))) {
-			ret = __try_to_free_cp_buf(jh);
-			if (ret) {
-				freed++;
-				if (ret == 2) {
-					*released = 1;
-					return freed;
-				}
+		ret = __try_to_free_cp_buf(jh);
+		if (ret) {
+			freed++;
+			if (ret == 2) {
+				*released = 1;
+				return freed;
 			}
 		}
 		/*
@@ -579,14 +582,13 @@ out:
  * checkpoint lists.
  *
  * The function returns 1 if it frees the transaction, 0 otherwise.
+ * The function can free jh and bh.
  *
- * This function is called with the journal locked.
  * This function is called with j_list_lock held.
- * This function is called with jbd_lock_bh_state(jh2bh(jh))
  */
-
 int __jbd2_journal_remove_checkpoint(struct journal_head *jh)
 {
+	struct transaction_chp_stats_s *stats;
 	transaction_t *transaction;
 	journal_t *journal;
 	int ret = 0;
@@ -599,13 +601,14 @@ int __jbd2_journal_remove_checkpoint(struct journal_head *jh)
 	}
 	journal = transaction->t_journal;
 
+	JBUFFER_TRACE(jh, "removing from transaction");
 	__buffer_unlink(jh);
 	jh->b_cp_transaction = NULL;
+	jbd2_journal_put_journal_head(jh);
 
 	if (transaction->t_checkpoint_list != NULL ||
 	    transaction->t_checkpoint_io_list != NULL)
 		goto out;
-	JBUFFER_TRACE(jh, "transaction has no more buffers");
 
 	/*
 	 * There is one special case to worry about: if we have just pulled the
@@ -616,22 +619,22 @@ int __jbd2_journal_remove_checkpoint(struct journal_head *jh)
 	 * The locking here around t_state is a bit sleazy.
 	 * See the comment at the end of jbd2_journal_commit_transaction().
 	 */
-	if (transaction->t_state != T_FINISHED) {
-		JBUFFER_TRACE(jh, "belongs to running/committing transaction");
+	if (transaction->t_state != T_FINISHED)
 		goto out;
-	}
 
 	/* OK, that was the last buffer for the transaction: we can now
 	   safely remove this transaction from the log */
+	stats = &transaction->t_chp_stats;
+	if (stats->cs_chp_time)
+		stats->cs_chp_time = jbd2_time_diff(stats->cs_chp_time,
+						    jiffies);
+	trace_jbd2_checkpoint_stats(journal->j_fs_dev->bd_dev,
+				    transaction->t_tid, stats);
 
 	__jbd2_journal_drop_transaction(journal, transaction);
-
-	/* Just in case anybody was waiting for more transactions to be
-           checkpointed... */
-	wake_up(&journal->j_wait_logspace);
+	jbd2_journal_free_transaction(transaction);
 	ret = 1;
 out:
-	JBUFFER_TRACE(jh, "exit");
 	return ret;
 }
 
@@ -650,6 +653,8 @@ void __jbd2_journal_insert_checkpoint(struct journal_head *jh,
 	J_ASSERT_JH(jh, buffer_dirty(jh2bh(jh)) || buffer_jbddirty(jh2bh(jh)));
 	J_ASSERT_JH(jh, jh->b_cp_transaction == NULL);
 
+	/* Get reference for checkpointing transaction */
+	jbd2_journal_grab_journal_head(jh2bh(jh));
 	jh->b_cp_transaction = transaction;
 
 	if (!transaction->t_checkpoint_list) {
@@ -688,17 +693,15 @@ void __jbd2_journal_drop_transaction(journal_t *journal, transaction_t *transact
 
 	J_ASSERT(transaction->t_state == T_FINISHED);
 	J_ASSERT(transaction->t_buffers == NULL);
-	J_ASSERT(transaction->t_sync_datalist == NULL);
 	J_ASSERT(transaction->t_forget == NULL);
-	J_ASSERT(transaction->t_iobuf_list == NULL);
 	J_ASSERT(transaction->t_shadow_list == NULL);
-	J_ASSERT(transaction->t_log_list == NULL);
 	J_ASSERT(transaction->t_checkpoint_list == NULL);
 	J_ASSERT(transaction->t_checkpoint_io_list == NULL);
-	J_ASSERT(transaction->t_updates == 0);
+	J_ASSERT(atomic_read(&transaction->t_updates) == 0);
 	J_ASSERT(journal->j_committing_transaction != transaction);
 	J_ASSERT(journal->j_running_transaction != transaction);
 
+	trace_jbd2_drop_transaction(journal, transaction);
+
 	jbd_debug(1, "Dropping transaction %d, all done\n", transaction->t_tid);
-	kfree(transaction);
 }