From: Theodore Ts'o Subject: Re: Question on resize2fs with "combination of flex_bg and !resize_inode" Date: Thu, 8 Nov 2012 18:03:08 -0500 Message-ID: <20121108230308.GN19977@thunk.org> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-ext4@vger.kernel.org To: Mark Casey Return-path: Received: from li9-11.members.linode.com ([67.18.176.11]:60129 "EHLO imap.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756742Ab2KHXDM (ORCPT ); Thu, 8 Nov 2012 18:03:12 -0500 Content-Disposition: inline In-Reply-To: Sender: linux-ext4-owner@vger.kernel.org List-ID: On Thu, Nov 08, 2012 at 02:03:38PM -0600, Mark Casey wrote: > Regarding this patch to resize2fs: > http://permalink.gmane.org/gmane.comp.file-systems.ext4/7379 > > In reading the comment, I was wondering if this problem scenario can > also happen when shrinking an ext4 filesystem with this combination > of features, or if it is only likely to be a problem when growing > it? It's only a problem when growing the file system. The comment in the patch explains the situation; when you are shrinking the file system, there is never any need to move inode tables to make room for the new block group descriptor blocks --- which are necessitated by the additional block groups that come from growing the file system. - Ted