Re: [PATCH 0/8][RFC] IO latency/throughput fixes

[Chris Mason <chris.mason@oracle.com>]

On Mon, 2009-04-06 at 08:37 -0700, Linus Torvalds wrote:
> 
>
> One thing that I find intriguing is how the fsync time seems so 
> _consistent_ across a wild variety of drives. It's interesting how you see 
> delays that are roughly the same order of magnitude, even though you are 
> using an old SATA drive, and I'm using the Intel SSD. And when you turn 
> off TCQ, your numbers go down even more.
> 

It may help if we rule out the ext3/jbd transaction growing code.  I'd
be pretty surprised if it made for long pauses, but its worth checking:

-chris

diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c
index e6a1174..a7c4f79 100644
--- a/fs/jbd/transaction.c
+++ b/fs/jbd/transaction.c
@@ -1417,7 +1417,7 @@ int journal_stop(handle_t *handle)
 	 * for joiners in that case.
 	 */
 	pid = current->pid;
-	if (handle->h_sync && journal->j_last_sync_writer != pid) {
+	if (0 && handle->h_sync && journal->j_last_sync_writer != pid) {
 		u64 commit_time, trans_time;
 
 		journal->j_last_sync_writer = pid;