linux-ext4 - Re: [PATCH 1/3] fscrypt: add support for inline-encryption-optimized policies

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20191024024459.GA743@sol.localdomain>
Date:   Wed, 23 Oct 2019 19:44:59 -0700
From:   Eric Biggers <ebiggers@...nel.org>
To:     Christoph Hellwig <hch@...radead.org>
Cc:     "Theodore Y. Ts'o" <tytso@....edu>,
        Satya Tangirala <satyat@...gle.com>,
        Paul Lawrence <paullawrence@...gle.com>,
        Dave Chinner <david@...morbit.com>,
        linux-f2fs-devel@...ts.sourceforge.net,
        linux-fscrypt@...r.kernel.org, linux-fsdevel@...r.kernel.org,
        Jaegeuk Kim <jaegeuk@...nel.org>, linux-ext4@...r.kernel.org,
        Paul Crowley <paulcrowley@...gle.com>
Subject: Re: [PATCH 1/3] fscrypt: add support for inline-encryption-optimized
 policies

On Wed, Oct 23, 2019 at 06:27:59PM -0700, Christoph Hellwig wrote:
> > If and when the vaporware shows up in real hardware, and assuming that
> > fscrypt is useful for this hardware, we can name it
> > "super_duper_fancy_inline_crypto".  :-)
> 
> I think you are entirely missing the point.  The point is that naming
> the option someting related to inline encryption is fundamentally
> wrong.  It is related to a limitation of existing inline crypto
> engines, not related to the fudamental model.  And all the other
> rambling below don't matter either.
> 

Would you be happy with something that more directly describes the change the
flag makes, like FSCRYPT_POLICY_FLAG_CONTENTS_IV_INO_LBLK_64?  I.e., the IVs for
contents encryption are 64-bit and contain the inode and logical block numbers.

Actually, we could use the same key derivation and IV generation for directories
and symlinks too, which would result in just FSCRYPT_POLICY_FLAG_IV_INO_LBLK_64.
(lblk is 0 when encrypting a filename.)

Although, in general it would be nice to name the settings in ways that are
easier for people not intimately familiar with the crypto to understand...

- Eric