3 files changed, 87 insertions, 15 deletions

diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 29148a81c783..7f21cf3aaf92 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -219,7 +219,6 @@ static int journal_submit_data_buffers(journal_t *journal,
                         ret = err;
                 spin_lock(&journal->j_list_lock);
                 J_ASSERT(jinode->i_transaction == commit_transaction);
-                commit_transaction->t_flushed_data_blocks = 1;
                 clear_bit(__JI_COMMIT_RUNNING, &jinode->i_flags);
                 smp_mb__after_clear_bit();
                 wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING);
@@ -672,12 +671,16 @@ start_journal_io:
                 err = 0;
         }
 
+        write_lock(&journal->j_state_lock);
+        J_ASSERT(commit_transaction->t_state == T_COMMIT);
+        commit_transaction->t_state = T_COMMIT_DFLUSH;
+        write_unlock(&journal->j_state_lock);
         /* 
          * If the journal is not located on the file system device,
          * then we must flush the file system device before we issue
          * the commit record
          */
-        if (commit_transaction->t_flushed_data_blocks &&
+        if (commit_transaction->t_need_data_flush &&
             (journal->j_fs_dev != journal->j_dev) &&
             (journal->j_flags & JBD2_BARRIER))
                 blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL);
@@ -754,8 +757,13 @@ wait_for_iobuf:
                    required. */
                 JBUFFER_TRACE(jh, "file as BJ_Forget");
                 jbd2_journal_file_buffer(jh, commit_transaction, BJ_Forget);
-                /* Wake up any transactions which were waiting for this
-                   IO to complete */
+                /*
+                 * Wake up any transactions which were waiting for this IO to
+                 * complete. The barrier must be here so that changes by
+                 * jbd2_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);
@@ -794,6 +802,10 @@ wait_for_iobuf:
                 jbd2_journal_abort(journal, err);
 
         jbd_debug(3, "JBD: commit phase 5\n");
+        write_lock(&journal->j_state_lock);
+        J_ASSERT(commit_transaction->t_state == T_COMMIT_DFLUSH);
+        commit_transaction->t_state = T_COMMIT_JFLUSH;
+        write_unlock(&journal->j_state_lock);
 
         if (!JBD2_HAS_INCOMPAT_FEATURE(journal,
                                        JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT)) {
@@ -949,7 +961,7 @@ restart_loop:
 
         jbd_debug(3, "JBD: commit phase 7\n");
 
-        J_ASSERT(commit_transaction->t_state == T_COMMIT);
+        J_ASSERT(commit_transaction->t_state == T_COMMIT_JFLUSH);
 
         commit_transaction->t_start = jiffies;
         stats.run.rs_logging = jbd2_time_diff(stats.run.rs_logging,
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index e0ec3db1c395..9a7826990304 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -479,9 +479,12 @@ int __jbd2_log_space_left(journal_t *journal)
 int __jbd2_log_start_commit(journal_t *journal, tid_t target)
 {
         /*
-         * Are we already doing a recent enough commit?
+         * The only transaction we can possibly wait upon is the
+         * currently running transaction (if it exists).  Otherwise,
+         * the target tid must be an old one.
          */
-        if (!tid_geq(journal->j_commit_request, target)) {
+        if (journal->j_running_transaction &&
+            journal->j_running_transaction->t_tid == target) {
                 /*
                  * We want a new commit: OK, mark the request and wakeup the
                  * commit thread.  We do _not_ do the commit ourselves.
@@ -493,7 +496,15 @@ int __jbd2_log_start_commit(journal_t *journal, tid_t target)
                           journal->j_commit_sequence);
                 wake_up(&journal->j_wait_commit);
                 return 1;
-        }
+        } else if (!tid_geq(journal->j_commit_request, target))
+                /* This should never happen, but if it does, preserve
+                   the evidence before kjournald goes into a loop and
+                   increments j_commit_sequence beyond all recognition. */
+                WARN_ONCE(1, "jbd: bad log_start_commit: %u %u %u %u\n",
+                          journal->j_commit_request,
+                          journal->j_commit_sequence,
+                          target, journal->j_running_transaction ? 
+                          journal->j_running_transaction->t_tid : 0);
         return 0;
 }
 
@@ -577,6 +588,47 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid)
 }
 
 /*
+ * Return 1 if a given transaction has not yet sent barrier request
+ * connected with a transaction commit. If 0 is returned, transaction
+ * may or may not have sent the barrier. Used to avoid sending barrier
+ * twice in common cases.
+ */
+int jbd2_trans_will_send_data_barrier(journal_t *journal, tid_t tid)
+{
+        int ret = 0;
+        transaction_t *commit_trans;
+
+        if (!(journal->j_flags & JBD2_BARRIER))
+                return 0;
+        read_lock(&journal->j_state_lock);
+        /* Transaction already committed? */
+        if (tid_geq(journal->j_commit_sequence, tid))
+                goto out;
+        commit_trans = journal->j_committing_transaction;
+        if (!commit_trans || commit_trans->t_tid != tid) {
+                ret = 1;
+                goto out;
+        }
+        /*
+         * Transaction is being committed and we already proceeded to
+         * submitting a flush to fs partition?
+         */
+        if (journal->j_fs_dev != journal->j_dev) {
+                if (!commit_trans->t_need_data_flush ||
+                    commit_trans->t_state >= T_COMMIT_DFLUSH)
+                        goto out;
+        } else {
+                if (commit_trans->t_state >= T_COMMIT_JFLUSH)
+                        goto out;
+        }
+        ret = 1;
+out:
+        read_unlock(&journal->j_state_lock);
+        return ret;
+}
+EXPORT_SYMBOL(jbd2_trans_will_send_data_barrier);
+
+/*
  * Wait for a specified commit to complete.
  * The caller may not hold the journal lock.
  */
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index 05fa77a23711..3eec82d32fd4 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -82,7 +82,7 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction)
  */
 
 /*
- * Update transiaction's maximum wait time, if debugging is enabled.
+ * Update transaction's maximum wait time, if debugging is enabled.
  *
  * In order for t_max_wait to be reliable, it must be protected by a
  * lock.  But doing so will mean that start_this_handle() can not be
@@ -91,11 +91,10 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction)
  * means that maximum wait time reported by the jbd2_run_stats
  * tracepoint will always be zero.
  */
-static inline void update_t_max_wait(transaction_t *transaction)
+static inline void update_t_max_wait(transaction_t *transaction,
+                                     unsigned long ts)
 {
 #ifdef CONFIG_JBD2_DEBUG
-        unsigned long ts = jiffies;
-
         if (jbd2_journal_enable_debug &&
             time_after(transaction->t_start, ts)) {
                 ts = jbd2_time_diff(ts, transaction->t_start);
@@ -121,6 +120,7 @@ static int start_this_handle(journal_t *journal, handle_t *handle,
         tid_t           tid;
         int             needed, need_to_start;
         int             nblocks = handle->h_buffer_credits;
+        unsigned long ts = jiffies;
 
         if (nblocks > journal->j_max_transaction_buffers) {
                 printk(KERN_ERR "JBD: %s wants too many credits (%d > %d)\n",
@@ -271,7 +271,7 @@ repeat:
         /* OK, account for the buffers that this operation expects to
          * use and add the handle to the running transaction. 
          */
-        update_t_max_wait(transaction);
+        update_t_max_wait(transaction, ts);
         handle->h_transaction = transaction;
         atomic_inc(&transaction->t_updates);
         atomic_inc(&transaction->t_handle_count);
@@ -316,7 +316,8 @@ static handle_t *new_handle(int nblocks)
  * This function is visible to journal users (like ext3fs), so is not
  * called with the journal already locked.
  *
- * Return a pointer to a newly allocated handle, or NULL on failure
+ * Return a pointer to a newly allocated handle, or an ERR_PTR() value
+ * on failure.
  */
 handle_t *jbd2__journal_start(journal_t *journal, int nblocks, int gfp_mask)
 {
@@ -921,8 +922,8 @@ int jbd2_journal_get_create_access(handle_t *handle, struct buffer_head *bh)
          */
         JBUFFER_TRACE(jh, "cancelling revoke");
         jbd2_journal_cancel_revoke(handle, jh);
-        jbd2_journal_put_journal_head(jh);
 out:
+        jbd2_journal_put_journal_head(jh);
         return err;
 }
 
@@ -2147,6 +2148,13 @@ int jbd2_journal_file_inode(handle_t *handle, struct jbd2_inode *jinode)
             jinode->i_next_transaction == transaction)
                 goto done;
 
+        /*
+         * We only ever set this variable to 1 so the test is safe. Since
+         * t_need_data_flush is likely to be set, we do the test to save some
+         * cacheline bouncing
+         */
+        if (!transaction->t_need_data_flush)
+                transaction->t_need_data_flush = 1;
         /* On some different transaction's list - should be
          * the committing one */
         if (jinode->i_transaction) {