Re: [PATCH v7 4/4] MAINTAINERS: Add entry for TEE based Trusted Keys

[Jarkko Sakkinen]

On Tue, Oct 13, 2020 at 04:58:47PM +0530, Sumit Garg wrote:
> On Tue, 13 Oct 2020 at 07:52, Jarkko Sakkinen
> <jarkko.sakkinen@xxxxxxxxxxxxxxx> wrote:
> >
> > On Wed, Oct 07, 2020 at 03:37:48PM +0530, Sumit Garg wrote:
> > > Add MAINTAINERS entry for TEE based Trusted Keys framework.
> > >
> > > Signed-off-by: Sumit Garg <sumit.garg@xxxxxxxxxx>
> > > Acked-by: Jarkko Sakkinen <jarkko.sakkinen@xxxxxxxxxxxxxxx>
> > > ---
> > >  MAINTAINERS | 8 ++++++++
> > >  1 file changed, 8 insertions(+)
> > >
> > > diff --git a/MAINTAINERS b/MAINTAINERS
> > > index 48aff80..eb3d889 100644
> > > --- a/MAINTAINERS
> > > +++ b/MAINTAINERS
> > > @@ -9663,6 +9663,14 @@ F:     include/keys/trusted-type.h
> > >  F:   include/keys/trusted_tpm.h
> > >  F:   security/keys/trusted-keys/
> > >
> > > +KEYS-TRUSTED-TEE
> > > +M:   Sumit Garg <sumit.garg@xxxxxxxxxx>
> > > +L:   linux-integrity@xxxxxxxxxxxxxxx
> > > +L:   keyrings@xxxxxxxxxxxxxxx
> > > +S:   Supported
> > > +F:   include/keys/trusted_tee.h
> > > +F:   security/keys/trusted-keys/trusted_tee.c
> > > +
> > >  KEYS/KEYRINGS
> > >  M:   David Howells <dhowells@xxxxxxxxxx>
> > >  M:   Jarkko Sakkinen <jarkko.sakkinen@xxxxxxxxxxxxxxx>
> > > --
> > > 2.7.4
> >
> > I'm sorry but I think I have changed my mind on this. This has been
> > spinning for a while and sometimes conclusions change over the time.
> >
> > I don't think that we really need a separate subsystem tag.
> 
> I don't see it as a separate subsystem but rather a kind of underlying
> trust source (TEE) driver plugged into existing trusted keys
> subsystem. We could relate it to the RNG subsystem as well where there
> is a subsystem maintainer and specific driver maintainers.
> 
> IMO, having a dedicated entry like this brings clarity in maintenance
> and in future we may have more trust sources like this added where
> everyone may not have access to all the trust sources to test.

More entries pointing to the exact same stuff does not necessarily mean
clarity in my books.

> > I'd be for a
> > new M-entry or R-entry to the existing subsystem tag. It's essential to
> > have ack from someone with ARM and TEE knowledge but this way too heavy
> > for the purpose.
> 
> If you still think otherwise then I am fine with a new M-entry for
> existing trusted keys subsystem as well.

Adding a M-entry does makes sense because trusted keys backends can be
based on various technologies and standard. It's a different in that
sense than lets say a TPM hardware driver.

> > I also see it the most manageable if the trusted keys PR's come from a
> > single source.
> 
> I echo here with you to have a single source for trusted keys PR's
> irrespective of whether we go with a separate trust source entry or
> update existing subsystem entry.
> 
> -Sumit

And I echo that oviously if there is someone to say the final ack about
TEE, I will require that as the minimum to ever pick any of those
changes :-)

I would resolve this with just the M-entry, and we can *later on*
restructure, if there is a need for that. These things are not sealed
to stone.

/Jarkko