summaryrefslogtreecommitdiff
path: root/fs/buffer.c
diff options
context:
space:
mode:
authorJens Axboe <axboe@suse.de>2006-06-13 08:26:10 +0200
committerJens Axboe <axboe@nelson.home.kernel.dk>2006-06-23 17:10:39 +0200
commitb31dc66a54ad986b6b73bdc49c8efc17cbad1833 (patch)
tree5591383c1cbffe11512da889c971f899333f1a44 /fs/buffer.c
parent271f18f102c789f59644bb6c53a69da1df72b2f4 (diff)
[PATCH] Kill PF_SYNCWRITE flag
A process flag to indicate whether we are doing sync io is incredibly ugly. It also causes performance problems when one does a lot of async io and then proceeds to sync it. Part of the io will go out as async, and the other part as sync. This causes a disconnect between the previously submitted io and the synced io. For io schedulers such as CFQ, this will cause us lost merges and suboptimal behaviour in scheduling. Remove PF_SYNCWRITE completely from the fsync/msync paths, and let the O_DIRECT path just directly indicate that the writes are sync by using WRITE_SYNC instead. Signed-off-by: Jens Axboe <axboe@suse.de>
Diffstat (limited to 'fs/buffer.c')
-rw-r--r--fs/buffer.c2
1 files changed, 0 insertions, 2 deletions
diff --git a/fs/buffer.c b/fs/buffer.c
index 23f1f3a68077..373bb6292bdc 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -331,7 +331,6 @@ long do_fsync(struct file *file, int datasync)
goto out;
}
- current->flags |= PF_SYNCWRITE;
ret = filemap_fdatawrite(mapping);
/*
@@ -346,7 +345,6 @@ long do_fsync(struct file *file, int datasync)
err = filemap_fdatawait(mapping);
if (!ret)
ret = err;
- current->flags &= ~PF_SYNCWRITE;
out:
return ret;
}