--u3/rZRmxL6MmkK24
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Fri, Oct 10, 2003 at 11:48:13AM -0700, Matt Thomas wrote:
> At 11:40 AM 10/10/2003, Bill Studenmund wrote:
> >This idea does mean we will have a transient allocation of locators, whi=
ch
> >are fed from the probe thread to the attach thread. For SPI, a packed
> >array of 2-bit ints may suffice (unprobed, nothing, just-LUN-0,
> >many-LUNs), but other busses may need something more tree-like. Once the=
se
> >locators get to the attach thread, they are turned into devices.
>=20
> Let's not make then transient.  Let's attach them permanently to the
> device so they can be obtained later.  We already keep the locator names,
> we just need to save the locator value (as an array of int) in the device
> struct.

We can copy them into the device struct later, but the reason I said=20
transient is that we don't have a struct device at probe time. :-)

Also it is not obvious that the best format for location info to tell a
bus what is on it is the same as the locators that tell a device what is=20
on it. Among other things, when you have the locator for a device, you=20
implicitly already know you have a device. :-)

Take care,

Bill

--u3/rZRmxL6MmkK24
Content-Type: application/pgp-signature
Content-Disposition: inline

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

iD8DBQE/itF3Wz+3JHUci9cRAlUMAJ0WPwpVmyykABLNXFE6cEh6u4BjxQCeKBCm
WqbjSI6yRKlFDZGTQvoUPbM=
=xoXp
-----END PGP SIGNATURE-----

--u3/rZRmxL6MmkK24--