Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
Date:   Thu, 14 Apr 2022 15:13:20 +0300
From:   Jarkko Sakkinen <jarkko@kernel.org>
To:     Jason Gunthorpe <jgg@ziepe.ca>
Cc:     Lukas Wunner <lukas@wunner.de>,
        Lino Sanfilippo <LinoSanfilippo@gmx.de>, peterhuewe@gmx.de,
        devicetree@vger.kernel.org, linux-integrity@vger.kernel.org,
        linux-kernel@vger.kernel.org, stefanb@linux.ibm.com,
        p.rosenberger@kunbus.com
Subject: Re: [PATCH 1/5] tpm: add functions to set and unset the tpm chips
 reset state
Message-ID: <YlgPxX3xCPUyR2F6@kernel.org>
References: <20220407111849.5676-1-LinoSanfilippo@gmx.de>
 <20220407111849.5676-2-LinoSanfilippo@gmx.de>
 <20220407142526.GW64706@ziepe.ca>
 <20220410171123.GA24453@wunner.de>
 <20220411114741.GA64706@ziepe.ca>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220411114741.GA64706@ziepe.ca>
Precedence: bulk

On Mon, Apr 11, 2022 at 08:47:41AM -0300, Jason Gunthorpe wrote:
> On Sun, Apr 10, 2022 at 07:11:23PM +0200, Lukas Wunner wrote:
> > On Thu, Apr 07, 2022 at 11:25:26AM -0300, Jason Gunthorpe wrote:
> > > On Thu, Apr 07, 2022 at 01:18:45PM +0200, Lino Sanfilippo wrote:
> > > > Currently it is not possible to set the tpm chips reset state from within
> > > > the driver. This is problematic if the chip is still in reset after the
> > > > system comes up. This may e.g. happen if the reset line is pulled into
> > > > reset state by a pin configuration in the device tree.
> > > 
> > > This kind of system is badly misdesigned.
> > > 
> > > TPM PCRs fundementally cannot work if the TPM reset line is under
> > > software control.
> > 
> > Not every system which incorporates a TPM wants to use or is even capable
> > of measuring software state of any kind or perform secure boot.
> > 
> > Those systems may merely want to use the TPM to store key material.
> 
> Then maybe the TPM driver should make it clear somehow that the PCRs
> don't work in these systems.
> 
> It is really dangerous to add capabilities like this that should
> never, ever be used in sanely designed systems.
> 
> Jason

I agree. That niche should do the bad things with oot patches.

BR, Jarkko