From: Dmitry Monakhov Subject: [PATCH 09/11] ext4: serialize truncate with owerwrite DIO workers V2 Date: Fri, 28 Sep 2012 19:44:09 +0400 Message-ID: <1348847051-6746-10-git-send-email-dmonakhov@openvz.org> References: <1348847051-6746-1-git-send-email-dmonakhov@openvz.org> Cc: tytso@mit.edu, jack@suse.cz, lczerner@redhat.com, Dmitry Monakhov To: linux-ext4@vger.kernel.org Return-path: Received: from mail-bk0-f46.google.com ([209.85.214.46]:49430 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758536Ab2I1Poc (ORCPT ); Fri, 28 Sep 2012 11:44:32 -0400 Received: by bkcjk13 with SMTP id jk13so3591839bkc.19 for ; Fri, 28 Sep 2012 08:44:31 -0700 (PDT) In-Reply-To: <1348847051-6746-1-git-send-email-dmonakhov@openvz.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: Jan Kara have spotted interesting issue: There are potential data corruption issue with direct IO overwrites racing with truncate: Like: dio write truncate_task ->ext4_ext_direct_IO ->overwrite == 1 ->down_read(&EXT4_I(inode)->i_data_sem); ->mutex_unlock(&inode->i_mutex); ->ext4_setattr() ->inode_dio_wait() ->truncate_setsize() ->ext4_truncate() ->down_write(&EXT4_I(inode)->i_data_sem); ->__blockdev_direct_IO ->ext4_get_block ->submit_io() ->up_read(&EXT4_I(inode)->i_data_sem); # truncate data blocks, allocate them to # other inode - bad stuff happens because # dio is still in flight. In order to serialize with truncate dio worker should grab extra i_dio_count reference before drop i_mutex. Changes agains V1: - wake up dio waiters before i_mutex. Reviewed-by: Jan Kara Signed-off-by: Dmitry Monakhov --- fs/ext4/inode.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index c40a98d..14b7096 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2982,6 +2982,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, overwrite = *((int *)iocb->private); if (overwrite) { + atomic_inc(&inode->i_dio_count); down_read(&EXT4_I(inode)->i_data_sem); mutex_unlock(&inode->i_mutex); } @@ -3079,6 +3080,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, retake_lock: /* take i_mutex locking again if we do a ovewrite dio */ if (overwrite) { + inode_dio_done(inode); up_read(&EXT4_I(inode)->i_data_sem); mutex_lock(&inode->i_mutex); } -- 1.7.7.6