From: Casey Schaufler Subject: Re: [RFC 0/28] Patches to pass vfsmount to LSM inode security hooks Date: Mon, 5 Feb 2007 12:23:24 -0800 (PST) Message-ID: <96672.70902.qm@web36613.mail.mud.yahoo.com> References: <20070205195058.GT10475@sequoia.sous-sol.org> Reply-To: casey@schaufler-ca.com Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT Cc: Tony Jones , linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, chrisw@sous-sol.org, linux-security-module@vger.kernel.org, agruen@suse.de To: Chris Wright Return-path: Received: from web36613.mail.mud.yahoo.com ([209.191.85.30]:45318 "HELO web36613.mail.mud.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S933365AbXBEUX0 (ORCPT ); Mon, 5 Feb 2007 15:23:26 -0500 In-Reply-To: <20070205195058.GT10475@sequoia.sous-sol.org> Sender: linux-fsdevel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org --- Chris Wright wrote: > It's not really worth describing, since it's not > acceptable in upstream. > But it basically cycles vfsmnts and looks for > matches to guess which > part of the tree the dentry is in. Yick. Indeed, that would be bad. > This kind of change (or perhaps > straight to struct path) is definitely > needed from AA. It is certainly tempting to suggest that the path struct scheme would be better. The only argument against it that I see is the kind of changes you'd have to make to existing code. It's not even a matter of having to make changes, just which changes you make. Casey Schaufler casey@schaufler-ca.com