From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: nobh mount option in ext3/4? Date: Mon, 31 May 2010 14:40:51 +0200 Message-ID: <20100531124051.GA17262@lst.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-ext4@vger.kernel.org, linux-fsdevel@vger.kernel.org To: jack@suse.cz, tytso@mit.edu Return-path: Received: from verein.lst.de ([213.95.11.210]:37686 "EHLO verein.lst.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754066Ab0EaMkz (ORCPT ); Mon, 31 May 2010 08:40:55 -0400 Content-Disposition: inline Sender: linux-ext4-owner@vger.kernel.org List-ID: ext3 and ext4 heave inherited the nobh option from ext2, but the implementation is rather interesting. Except for the usual option parsing / display it's only respected in ext3_writeback_writepage/ext4_writepage and the truncate_page implementation. Given that we don't take any more than usual care to not create buffers heads during read this is not going to effect a whole lot. Is there a good reason to keep these options and the associated code complexity? I'd be rather surprised if it gets any sufficient testing..