--yZnyZsPjQYjG7xG7
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Dec 11, 2003 at 09:28:43AM -0800, Jason Thorpe wrote:
>=20
> On Dec 11, 2003, at 9:21 AM, Bill Studenmund wrote:
>=20
> >Mostly. That paragraph was before I read the spec.
> >
> >The question that remains though came up later in the note. :-) How do=
=20
> >you
> >make sure the PCI->LPC bridge chip has been configured before you try=20
> >to
> >configure the things behind it?
>=20
> Right now, we assume the BIOS has done it.  We have assumed this for=20
> years.  A PCI->LPC bridge is just a fancy word for "southbridge that=20
> doesn't have ISA slot support".  All those southbridge and PCI superio=20
> chips are "PCI->LPC bridges".

I was afraid you'd say that. I'm not sure how much it's needed, but the=20
tcpa driver seems to need to explicitly enable the PCI->LCP bridge, which=
=20
would imply it's not done by the BIOS.

> >The LPC spec implies that devices on it should use ACPI for=20
> >configuration,
> >so the question may turn into how does ACPI tell us to configure the=20
> >LPC
> >bridge, and how do we ensure that the LPC bridge is configured before=20
> >the
> >(ACPI) devices behind it try to probe?
>=20
> I suggest you read the ACPI spec.  Keep it in your bathroom.

I was afraid you'd suggest that. I downloaded it, and there was only one=20
occurence of LPC in the spec, and it didn't talk about configuring an=20
PCI<->LPC bridge.

Take care,

Bill

--yZnyZsPjQYjG7xG7
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (NetBSD)

iD8DBQE/2LUwWz+3JHUci9cRAn18AJ4rZy/2qatZ6J7Jk0LWIra5ud9oYgCggLEl
8C2F5jNkKoQFlUXK+zPiHH8=
=7AxH
-----END PGP SIGNATURE-----

--yZnyZsPjQYjG7xG7--