From: Theodore Ts'o Subject: Re: An idea for supporting large directories and readdir+stat workloads Date: Mon, 13 Aug 2012 10:39:41 -0400 Message-ID: <20120813143941.GC32484@thunk.org> References: <20120811212532.GB906@thunk.org> <20120813143607.GB32484@thunk.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Jeff Moyer , Ext4 Developers List To: Andreas Dilger Return-path: Received: from li9-11.members.linode.com ([67.18.176.11]:44414 "EHLO imap.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751046Ab2HMOjn (ORCPT ); Mon, 13 Aug 2012 10:39:43 -0400 Content-Disposition: inline In-Reply-To: <20120813143607.GB32484@thunk.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Mon, Aug 13, 2012 at 10:36:07AM -0400, Theodore Ts'o wrote: > Field Size > > ino 4 Silly me, we don't need actually need to store the inode a second time, since that's already in the struct ext4_dir_entry. > dev 4 ... and we only need to store the device if this is a block or character device, so we could actually use the i_size field below to store the dev_t for devices inodes. So this would shrink the size of the "compact inode" that would be stored in the directory entry down from 62 bytes to 58 bytes. - Ted