[JBD] change batching logic to improve O_SYNC performance

[Benjamin LaHaise <bcrl@kvack.org>]

Hello folks,

When writing files out using O_SYNC, jbd's 1 jiffy delay results in a 
significant drop in throughput as the disk sits idle.  The patch below 
results in a 4-5x performance improvement (from 6.5MB/s to ~24-30MB/s on 
my IDE test box) when writing out files using O_SYNC.  Instead of always 
delaying for 1 jiffy when trying to batch, merely do a yield() to allow 
other processes to execute and potentially batch requests.  Additionally, 
limit the delay to an absolute max of 1/50th of a second.  

		-ben

Signed-off-by: Benjamin LaHaise <benjamin.c.lahaise@intel.com>
diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c
index 429f4b2..9c84ad0 100644
--- a/fs/jbd/transaction.c
+++ b/fs/jbd/transaction.c
@@ -1335,10 +1335,12 @@ int journal_stop(handle_t *handle)
 	 * by 30x or more...
 	 */
 	if (handle->h_sync) {
+		long start = jiffies;
 		do {
 			old_handle_count = transaction->t_handle_count;
-			schedule_timeout_uninterruptible(1);
-		} while (old_handle_count != transaction->t_handle_count);
+			yield();
+		} while (old_handle_count != transaction->t_handle_count &&
+			 (jiffies - start < HZ/50));
 	}
 
 	current->journal_info = NULL;
-- 
"You know, I've seen some crystals do some pretty trippy shit, man."
Don't Email: <dont@kvack.org>.