From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mandeep Singh Baines Subject: Re: [PATCH] dm: verity target Date: Thu, 29 Sep 2011 10:31:03 -0700 Message-ID: <20110929173103.GA22515@google.com> References: <1316112359-6085-1-git-send-email-msb@chromium.org> <21505.1317245414@turing-police.cc.vt.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <21505.1317245414@turing-police.cc.vt.edu> Sender: linux-kernel-owner@vger.kernel.org To: Valdis.Kletnieks@vt.edu Cc: Will Drewry , Mandeep Singh Baines , Alasdair G Kergon , Milan Broz , dm-devel@redhat.com, Elly Jones , Olof Johansson , linux-kernel@vger.kernel.org List-Id: dm-devel.ids Valdis.Kletnieks@vt.edu (Valdis.Kletnieks@vt.edu) wrote: > On Tue, 27 Sep 2011 14:02:05 CDT, Will Drewry said: > > > I was just curious if there is any interest in pulling this change, or > > if not, if there is any particular set of concerns, fixes, etc. > > Out of curiosity, how much of the stack does this end up eating? My root > filesystem is already ext4 on an LVM partition that's on a LUKS/dm-crypt > partition on a hard drive, and I'm sure somebody out there will have used xfs > instead - and then exported it via NFS or something. Are we going to get weird > stack overflows if people throw dm-verity into this sort of mix? > No. dm-verity uses very little stack since most of the code is running in a separate workqueue context. The _map call is pretty light. > > realize it's not a small amount of code to digest (though it is > > smaller than the post from last year[1]). Would re-posting with an > > added blob explaining the name be useful, > > Probably will need it to be merged, unless you set up an auto-reply that says > "Patch rejected, 'verity' is *not* a typo for 'verify'" ;) > > I'll hopefully have some more comments over the weekend if I get some spare > cycles. >