From: Jan Kara Subject: [PATCH 2/2] ext4: Fix waiting and sending of a barrier in ext4_sync_file() Date: Tue, 17 May 2011 12:28:15 +0200 Message-ID: <1305628095-27843-3-git-send-email-jack@suse.cz> References: <1305628095-27843-1-git-send-email-jack@suse.cz> Cc: Edward Goggin , linux-ext4@vger.kernel.org, Jan Kara To: tytso@mit.edu Return-path: Received: from cantor.suse.de ([195.135.220.2]:59690 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752836Ab1EQK2u (ORCPT ); Tue, 17 May 2011 06:28:50 -0400 In-Reply-To: <1305628095-27843-1-git-send-email-jack@suse.cz> Sender: linux-ext4-owner@vger.kernel.org List-ID: jbd2_log_start_commit() returns 1 only when we really start a transaction. But we also need to wait for a transaction when the commit is already running. Fix this problem by waiting for transaction commit unconditionally (which is just a quick check if the transaction is already committed). Also we have to be more careful with sending of a barrier because when transaction is being committed in parallel to ext4_sync_file() running, we cannot be sure that the barrier the journalling code sends happens after we wrote all the data for fsync (note that not every data writeout needs to trigger metadata changes thus commit of some metadata changes can be running while other data is still written out). So use jbd2_will_send_data_barrier() helper to detect the common cases when we can be sure barrier will be issued by the commit code and issue the barrier ourselves in the remaining cases. This code change will cause ext4_sync_file() to unnecessarily issue a barrier in data=ordered and data=journal mode with external journal but solving this would require ext4 to mess with jbd2 internals since jbd2 does not know about ext4 data journalling modes. Signed-off-by: Jan Kara --- fs/ext4/fsync.c | 23 +++++++---------------- 1 files changed, 7 insertions(+), 16 deletions(-) diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index e9473cb..1c44f1a 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -172,6 +172,7 @@ int ext4_sync_file(struct file *file, int datasync) journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; int ret; tid_t commit_tid; + bool needs_barrier = false; J_ASSERT(ext4_journal_current_handle() == NULL); @@ -211,22 +212,12 @@ int ext4_sync_file(struct file *file, int datasync) } commit_tid = datasync ? ei->i_datasync_tid : ei->i_sync_tid; - if (jbd2_log_start_commit(journal, commit_tid)) { - /* - * When the journal is on a different device than the - * fs data disk, we need to issue the barrier in - * writeback mode. (In ordered mode, the jbd2 layer - * will take care of issuing the barrier. In - * data=journal, all of the data blocks are written to - * the journal device.) - */ - if (ext4_should_writeback_data(inode) && - (journal->j_fs_dev != journal->j_dev) && - (journal->j_flags & JBD2_BARRIER)) - blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, - NULL); - ret = jbd2_log_wait_commit(journal, commit_tid); - } else if (journal->j_flags & JBD2_BARRIER) + if (journal->j_flags & JBD2_BARRIER && + !jbd2_trans_will_send_data_barrier(journal, commit_tid)) + needs_barrier = true; + jbd2_log_start_commit(journal, commit_tid); + ret = jbd2_log_wait_commit(journal, commit_tid); + if (needs_barrier) blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); out: trace_ext4_sync_file_exit(inode, ret); -- 1.7.1