[ltp] Re: Undocking?

[Alex Deucher]

On 8/23/06, Stefan Schmidt <stefan@sostec.de> wrote:
> Hello.
>
> On Thu, 2006-08-24 at 00:34, Shem Multinymous wrote:
> > On 8/24/06, Stefan Schmidt <stefan@sostec.de> wrote:
> > >> If that works, maybe it will give us a hint on how to solve the
> > >> analogous problem with the UltraBay: hotswapping using ibm_acpi works
> > >> only if you had a device inserted during boot.
> > >
> > >Right. The new docking driver lives inside the acpi, just like
> > >lt_hotswap. Perhaps we get a chance to merge the necessary bits form
> > >lt_hotswap in a clean way into this acpi-pci-hotplug-fiddling stuff.
> >
> > I think the problem starts at a lower level than t_hotswap: if there
> > wasn't an UltraBay device inserted during boot, then
> > /proc/acpi/ibm/bay isn't created and no ACPI events are generated on
> > UltraBay insertion/removal.
>
> Yes, but the new docking stuff do not base on ibm-acpi. As they handle
> the acpi and pci-hotplugging stuff on their own, i hope we don't run
> into this limitation again. ACPI should really know about a new
> inserted device, as windows can handle this.
>
> But i'm not firm with the ibm-acpi code either with the new docking
> driver so perhaps i_m wrong.

ibm_acpi is going to need some work to interact properly with the new
acpi docking code.  right now they both try and claim/use the same
acpi parts.

Alex

>
> regards
> Stefan Schmidt
>
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.1 (GNU/Linux)
> Comment: http://www.datenfreihafen.org/contact.html
>
> iD8DBQFE7MoabNSsvd31FmURAn+0AJ9yONuB58bjD6bYtTxnsdw0eSKeDwCdFgHS
> 8eOCbHGF4zFIGdEb3V3qcM8=
> =v2QP
> -----END PGP SIGNATURE-----
>
>
>