From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mike Snitzer Subject: Re: dm-crypt: Reject sector_size feature if device length is not aligned to it Date: Tue, 3 Oct 2017 14:08:04 -0400 Message-ID: <20171003180804.GA25465@redhat.com> References: <20170913134556.23145-1-gmazyland@gmail.com> <20171003120508.GA9979@agk-dp.fab.redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20171003120508.GA9979@agk-dp.fab.redhat.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: Alasdair G Kergon Cc: dm-devel@redhat.com, Mikulas Patocka , Milan Broz List-Id: dm-devel.ids On Tue, Oct 03 2017 at 8:05am -0400, Alasdair G Kergon wrote: > On Mon, Oct 02, 2017 at 10:43:56AM -0400, Mikulas Patocka wrote: > > Here I send a patch that checks invalid limits when the table is loaded > > and aborts the table load ioctl with an error. > > The code does not already do this because it needs to avoid unnecessary > failures if any characteristics of the devices underneath change after > the 'load' but before the 'resume' (as can happen, for example, when a > tree of devices gets updated by performing all the loads before all the > resumes). Not sure why you're putting such value on that behaviour. The earlier we catch invalid tables the better off we are. Failing at resume time sucks (always has). Could be there is some specific scenario where lvm2 is leveraging this historic DM "feature" but it'd be nice to revalidate the need. Mike