| To: | Matthias Schniedermeyer <ms@xxxxxxx> |
|---|---|
| Subject: | Re: XFS doesn't correctly account for IO-Wait for directory reading |
| From: | Christoph Hellwig <hch@xxxxxxxxxxxxx> |
| Date: | Wed, 23 Jan 2008 12:17:41 +0000 |
| Cc: | xfs@xxxxxxxxxxx |
| In-reply-to: | <20080123110027.GA10366@citd.de> |
| References: | <20080123110027.GA10366@citd.de> |
| Sender: | xfs-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.5.17 (2007-11-01) |
Try this one-liner patch which should give you much better I/O wait
reporting. There's some more I/O waits hidden in the log code, but to
fix this we'd need to dig into the sv_t abstraction. Given that it only
has four users left I'm probably going to simply remove it and fix the
I/O wait accounting while I'm at it.
--- linux-2.6-xfs.orig/fs/xfs/linux-2.6/xfs_buf.c 2008-01-23
13:08:48.000000000 +0100
+++ linux-2.6-xfs/fs/xfs/linux-2.6/xfs_buf.c 2008-01-23 13:08:54.000000000
+0100
@@ -976,7 +976,7 @@ xfs_buf_wait_unpin(
break;
if (atomic_read(&bp->b_io_remaining))
blk_run_address_space(bp->b_target->bt_mapping);
- schedule();
+ io_schedule();
}
remove_wait_queue(&bp->b_waiters, &wait);
set_current_state(TASK_RUNNING);
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | XFS doesn't correctly account for IO-Wait for directory reading, Matthias Schniedermeyer |
|---|---|
| Next by Date: | Re: blktrace & btrace usability, Jan Blunck |
| Previous by Thread: | XFS doesn't correctly account for IO-Wait for directory reading, Matthias Schniedermeyer |
| Next by Thread: | Re: XFS doesn't correctly account for IO-Wait for directory reading, Matthias Schniedermeyer |
| Indexes: | [Date] [Thread] [Top] [All Lists] |