RE: [RFC PATCH 4/5] crypto: Adds user space interface for ALG_SET_KEY_TYPE

[Kalyani Akula]

Ping!!

> -----Original Message-----
> From: Kalyani Akula
> Sent: Wednesday, May 8, 2019 3:01 PM
> To: Stephan Mueller <smueller@xxxxxxxxxx>
> Cc: herbert@xxxxxxxxxxxxxxxxxxx; davem@xxxxxxxxxxxxx; linux-
> crypto@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; Sarat Chand Savitala
> <saratcha@xxxxxxxxxx>
> Subject: RE: [RFC PATCH 4/5] crypto: Adds user space interface for
> ALG_SET_KEY_TYPE
> 
> Hi Stephan,
> 
> Keyrings is in-kernel key-management and retention facility. User can use it to
> manage keys used for applications.
> 
> Xilinx cryptographic hardware has a mechanism to store keys in its internal
> hardware and do not have mechanism to read it back due to security reasons.
> It stores key internally in different forms like simple key, key encrypted with
> unique hardware DNA, key encrypted with hardware family key, key stored in
> eFUSEs/BBRAM etc.
> Based on security level expected, user can select one of the key for AES
> operation. Since AES hardware internally has access to these keys, user do not
> require to provide key to hardware, but need to tell which internal hardware key
> user application like to use for AES operation.
> 
> Basic need is to pass information to AES hardware about which internal
> hardware key to be used for AES operation.
> 
> I agree that from general use case perspective we are not selecting key type but
> selecting internal hardware keys provided by user.
> How about providing option to select custom hardware keys provided by
> hardware (AES_SEL_HW_KEY)?
> 
> Thanks
> kalyani
> 
> > -----Original Message-----
> > From: Stephan Mueller <smueller@xxxxxxxxxx>
> > Sent: Thursday, April 25, 2019 12:01 AM
> > To: Kalyani Akula <kalyania@xxxxxxxxxx>
> > Cc: herbert@xxxxxxxxxxxxxxxxxxx; davem@xxxxxxxxxxxxx; linux-
> > crypto@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
> > Subject: Re: [RFC PATCH 4/5] crypto: Adds user space interface for
> > ALG_SET_KEY_TYPE
> >
> > Am Montag, 22. April 2019, 11:17:55 CEST schrieb Kalyani Akula:
> >
> > Hi Kalyani,
> >
> > > > Besides, seem to be more a key handling issue. Wouldn't it make
> > > > sense to rather have such issue solved with key rings than in the
> > > > kernel crypto API?
> > >
> > > [kalyani] Can you please elaborate on this further ?
> >
> > The kernel has a keyring support in security/keys which has a user
> > space interface with keyutils. That interface is commonly used for any
> > sort of key manipulation.
> >
> > Ciao
> > Stephan
> >