From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH-v4 1/7] vfs: split update_time() into update_time() and write_time() Date: Thu, 27 Nov 2014 07:25:04 -0800 Message-ID: <20141127152504.GA15839@infradead.org> References: <1416997437-26092-1-git-send-email-tytso@mit.edu> <1416997437-26092-2-git-send-email-tytso@mit.edu> <20141126192328.GA20436@infradead.org> <20141127123429.GD30152@quack.suse.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Theodore Ts'o , Linux Filesystem Development List , Ext4 Developers List , Linux btrfs Developers List , XFS Developers To: Jan Kara Return-path: Received: from bombadil.infradead.org ([198.137.202.9]:46071 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750781AbaK0PZG (ORCPT ); Thu, 27 Nov 2014 10:25:06 -0500 Content-Disposition: inline In-Reply-To: <20141127123429.GD30152@quack.suse.cz> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Thu, Nov 27, 2014 at 01:34:29PM +0100, Jan Kara wrote: > But Ted changed XFS to copy timestamps to on-disk structure from the > in-memory inode fields after VFS updated the timestamps. So the stamps > should be coherent AFAICT, shouldn't they? Not coherent enough. We need the XFS ilock to synchronize reading from and writing to the time stamps. update_time() only has i_mutex, which we can't take for the transaction commit path.