From: Chris Mason <chris.mason@oracle.com>
Date: Tue, 27 May 2008 14:52:17 +0000 (-0400)
Subject: Btrfs: Always use the async submission queue for checksummed writes
X-Git-Tag: v2.6.29-rc2~3^2~11^2~55^2~56^2~30^2~165
X-Git-Url: http://ftp.safe.ca/?a=commitdiff_plain;h=da496f2acf61153e0d4d42ded6a53f3367939db0;p=safe%2Fjmp%2Flinux-2.6

Btrfs: Always use the async submission queue for checksummed writes

This avoids IO stalls and poorly ordered IO from inline writers mixing in
with the async submission queue

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index d8a256d..016c0c0 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -434,13 +434,6 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
 {
 	struct async_submit_bio *async;
 
-	/*
-	 * inline writerback should stay inline, only hop to the async
-	 * queue if we're pdflush
-	 */
-	if (!current_is_pdflush())
-		return submit_bio_hook(inode, rw, bio, mirror_num);
-
 	async = kmalloc(sizeof(*async), GFP_NOFS);
 	if (!async)
 		return -ENOMEM;