Re: [RFC/PATCH 4/4] Add support for MSI on Axon-based Cell systems

[Arnd Bergmann <arnd@arndb.de>]

On Tuesday 05 June 2007, Jean-Christophe Dubois wrote:
> On Monday 04 June 2007 15:00:05 Michael Ellerman wrote:
> > +=A0=A0=A0=A0=A0=A0=A0pr_debug("axon_msi: initialising ...\n");
> > +
> > +=A0=A0=A0=A0=A0=A0=A0for_each_compatible_node(node, NULL, "ibm,axon-ms=
ic") {
> > +=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0if (axon_msi_setup_one(of=
_node_get(node)) =3D=3D 0)
> > +=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0f=
ound++;
> > +=A0=A0=A0=A0=A0=A0=A0}
> > +=A0=A0=A0=A0=A0=A0=A0of_node_put(node);
>=20
> 2 comments:
>=20
> 1) There is no "ibm,axon-msic" compatible property in the SLOF tree provi=
ded=20
> by IBM on the CAB (or other such platforms). Therefore this code will not=
=20
> work on these platforms.

Right, the firmware needs to support msi so we can use it. I guess taht's t=
rue
for every device: If the firmware decies to hide it from us, we better don't
try to touch it.

> 2) you should somehow check for the Axon version. Axon 1.1 (and prior) ar=
e=20
> know to be loosing MSI/MBX interrupts (but not the data part associated t=
o=20
> them). Therefore this should be enabled only on Axon 2.1 (or later).

Good point, thanks for mentioning it. Christian, can you confirm that the
ibm,axon-msic node is only created on systems with axon 2.1 or later?

	Arnd <><