Re: [tpmdd-devel] Last thing for force=1 rework

[Jason G. <jgu...@ob...>]

On Thu, Dec 10, 2015 at 11:40:23AM +0200, Jarkko Sakkinen wrote:
> On Wed, Dec 09, 2015 at 02:49:55PM -0700, Jason Gunthorpe wrote:
> > On Wed, Dec 09, 2015 at 11:45:40PM +0200, Jarkko Sakkinen wrote:
> > 
> > > When I modprobe tpm_tis force=1 I get:
> > > 
> > > [ 2547.438038] tpm_tis tpm_tis: 2.0 TPM (device-id 0x1A, rev-id 16)
> > > [ 2547.750326] tpm_tis MSFT0101:00: can't request region for resource
> > > [mem 0xfed40000-0xfed44fff]
> > > [ 2547.750334] tpm_tis: probe of MSFT0101:00 failed with error -16
> > 
> > This is expected and fine, I'll add your Tested-by for the tis stuff
> 
> Sorry to ask a stupid question but why this is expected and fine?

The new version for force doesn't inhibit auto detection, so the
force'd device attaches first:

> > > [ 2547.438038] tpm_tis tpm_tis: 2.0 TPM (device-id 0x1A, rev-id 16)

Then auto probing runs, and is blocked because the force driver took
over the address range first:

> > > [ 2547.750326] tpm_tis MSFT0101:00: can't request region for resource
> > > [mem 0xfed40000-0xfed44fff]

Using force when pnp is already working for the device is a user
mistake and something worth logging.

Jason