Re: [PATCH] cipher: add AES-CMAC hashing support

[Gix, Brian <brian.gix@intel.com>]

[-- Attachment #1: Type: text/plain, Size: 1705 bytes --]

Hi Michal,


On Wed, 2019-05-29 at 21:26 +0200, michal.lowas-rzechonek(a)silvair.com wrote:
> Hi Brian, Denis,
> 
> On 05/29, Gix, Brian wrote:
> > > Is this about mesh_crypto_aes_ccm_encrypt and
> > > mesh_crypto_aes_ccm_decrypt functions in mesh/crypto.c?
> > 
> > The code you copy pasted below *is* the usage of AES_CCM that we use in mesh,
> 
> Um, it's not - mesh_crypto_aes_ccm_* functions seem to implement AES_CCM
> on top of ECB, in user space. As Denis mentioned, this is somewhat
> inefficient, but at least it works on older kernels.

Yes, I understand now...  You wrote the code snippet, that looked an awful lot like the patch I was in the
middle of writing for mesh_crypto_aes_ccm_*....

my best guess right now is that we will probably decide to go ahead and patch mesh such that it assumes an up-
to-date kernel version, and a working underlying architecture.

It will then be the responsibility of the vendor (you guys) to patch mesh such that it continues to work on
your older kernels and perhaps architectures.  That way future kernels and platforms are not perpetually "held
back" to support kernels that should be obsoleted or fixed. 

> 
> I was thinking about submitting a patch to change that, but then I'd be
> shooting myself in the foot, because my target platform runs kernel 4.4
> if I recall correctly ;)
> 
> > but it is *not* the topic of the patch I posted here to ELL yesterday.
> > AES-CMAC is the algorithm we use in Mesh to generate all of the
> > various Keys and IDs from master Net and App keys....
> 
> Ah, right, it's the K* family of functions. Sorry, I confused CCM and
> CMAC modes.
> 
> regards

Best Reagrds,
Brian