From mboxrd@z Thu Jan 1 00:00:00 1970 From: Theodore Tso Subject: Re: [PATCH] allow goal inode to be specified Date: Sat, 13 Jun 2009 11:48:11 -0400 Message-ID: <20090613154811.GL24336@mit.edu> References: <20090610202702.GI9002@webber.adilger.int> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-ext4@vger.kernel.org To: Andreas Dilger Return-path: Received: from thunk.org ([69.25.196.29]:44474 "EHLO thunker.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753003AbZFMPsO (ORCPT ); Sat, 13 Jun 2009 11:48:14 -0400 Content-Disposition: inline In-Reply-To: <20090610202702.GI9002@webber.adilger.int> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Wed, Jun 10, 2009 at 02:27:02PM -0600, Andreas Dilger wrote: > Allow a goal inode to be specified when allocating new inodes. > If the goal is within the range of valid inodes use this in > preference to one of the other target inode heuristics (Orlov or > parent directory). > > The goal inode can be specified via /sys/fs/{dev}/inode_goal for > testing inode allocation beyond 2^32 blocks on very large filesystems, > and in the future will be used for large xattr inode allocation. This patch didn't quite apply against the mainline ext4, and in addition it conflicted with a patch to keep the choice of Orlov topdir inode numbers stable for the purposes of benchmarking. So instead I dropped the following two patches into the ext4 patch queue. - Ted