From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Biggers Date: Wed, 20 Feb 2019 18:36:20 +0000 Subject: Re: [RFC PATCH v3 00/18] fscrypt: key management improvements Message-Id: <20190220183619.GA177939@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: quoted-printable List-Id: References: <20190220065249.32099-1-ebiggers@kernel.org> <14023.1550686042@warthog.procyon.org.uk> In-Reply-To: <14023.1550686042@warthog.procyon.org.uk> To: David Howells Cc: linux-ext4@vger.kernel.org, linux-api@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net, linux-fscrypt@vger.kernel.org, keyrings@vger.kernel.org, linux-mtd@lists.infradead.org, linux-crypto@vger.kernel.org, linux-fsdevel@vger.kernel.org, Satya Tangirala , Paul Crowley Hi David, On Wed, Feb 20, 2019 at 06:07:22PM +0000, David Howells wrote: > I have a couple of patches that add ACLs to keyrings, that you can find a= t the > top of the branch here: >=20 > https://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs.git/lo= g/?h=3Dkeys-acl >=20 > I have other patches that allow tags to be used as subjects in the ACL, w= ith a > container supplying a tag, e.g.: >=20 > https://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs.git/co= mmit/?h=3Dcontainer&id=E0fdc5613a32b9475b025f58e7a2267329b3f3a4 >=20 > Might something similar be of use to you here, perhaps specifying a tag > referring to the blockdev of interest rather than the container? >=20 > David I don't think so. The main point of adding keys directly to the filesystem= is that it's generally inappropriate to apply OS-level access control or process-based visibility restrictions to fscrypt keys given that: - The unlocked/locked status of files is already filesystem-level. - The purpose of encryption is orthogonal to OS-level access control. The ioctl based interface also makes it *much* easier to implement all the semantics needed for removing fscrypt keys. I don't see how key ACLs would be appropriate here, except that key ACLs wo= uld allow userspace to opt-in to fixing the denial of service vulnerability whe= re keyctl_invalidate() only requires Search permission. But for fscrypt that's addressed by my proposal too, and is just one of many problems it addresses. - Eric