Re: [PATCH] tpm, tpm_tis: fix TPM 2.0 probing

[Jarkko Sakkinen <jarkko.sakkinen-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>]

On Wed, Feb 11, 2015 at 01:47:01PM -0500, Stefan Berger wrote:
> On 02/10/2015 07:50 AM, Jarkko Sakkinen wrote:
> >On Tue, Feb 10, 2015 at 07:16:32AM -0500, Stefan Berger wrote:
> >>On 02/09/2015 03:39 AM, Jarkko Sakkinen wrote:
> >>>On Mon, Feb 09, 2015 at 12:08:46AM +0100, Peter Hüwe wrote:
> >>>>Am Mittwoch, 4. Februar 2015, 15:21:09 schrieb Jarkko Sakkinen:
> >>>>>If during transmission system error was returned, the logic was to
> >>>>>incorrectly deduce that chip is a TPM 1.x chip. This patch fixes this
> >>>>>issue. Also, this patch changes probing so that message tag is used as the
> >>>>>measure for TPM 2.x, which should be much more stable.
> >>>>Is it aware that some TPMs may respond with 0x00C1 as TAG for TPM1.2 commands?
> >>>I guess none of the TPM 1.2 command answer with the tag 0x8002?
> >>
> >>FYI: pdf page 26 , section 6.1 explains the predictable return value for a
> >>TPM1.2 command seen by a TPM2
> >>
> >>http://www.trustedcomputinggroup.org/files/static_page_files/8C68ADA8-1A4B-B294-D0FC06D3773F7DAA/TPM%20Rev%202.0%20Part%203%20-%20Commands%2001.16-code.pdf
> >>
> >>Following this:
> >>
> >>Sending a TPM1.2 command to a TPM2 should return a TPM1.2 header (tag =
> >>0xc4) and error code (TPM_BADTAG = 0x1e)
> >>
> >>Sending a TPM 2 command to a TPM 2 will give a TPM 2 tag in the header.
> >>Sending a TPM 2 command to a TPM 1.2 will give a TPM 1.2 tag in the header
> >>and an error code.
> >Thank you for the information. Do you think that for some reason
> >tpm2_probe() shoould instead check that value is not this error
> >instead of checking that tag is 0x80002?
> 
> Following your path, you are checking for TPM2_ST_NO_SESSION (0x8001), which
> looks correct to me. A TPM1.2 would never send this tag back.

OK, perfect :)

>     Stefan

/Jarkko