PERFORCE change 36551 for review
John Baldwin
jhb at FreeBSD.org
Mon Aug 25 14:02:08 PDT 2003
On 22-Aug-2003 M. Warner Losh wrote:
> In message: <20030822173606.GA849 at dhcp42.pn.xcllnt.net>
> Marcel Moolenaar <marcel at xcllnt.net> writes:
>: > > So, please. Do not blur the distinction by having it all mapped as
>: > > ISA devices. I really don't want to have to shoot you :-)
>: >
>: > Well, the other alternative is to add an ACPI attachment for every
>: > ISA device. I'm sure you can appreciate my lack of zeal for this
>: > option. :(
>:
>: Moving forward I would think that you replace ISA bus attachments
>: with ACPI bus attachments until such time you don't have any ISA
>: drivers anymore. That is, you probably need drivers for ISA devices
>: that exist on alpha, so you'll end up with a handful of drivers
>: that have both ACPI and ISA. Hmmm, I don't know if this holds for
>: pc98 or not?
>
> Assumption: bus attachement are cheap.
>
> There will be a number of drivers that violate this, and may need to
> be rearranged, but forcing a device onto the ISA bus because we don't
> have a ACPI attachment for it is lame.
>
> The pc98 folks have expressed a strong desire that CBUS front ends be
> separate from ISA front ends. There's a number of technical reasons
> for this, in addition to the asthetic argument. Basically, they've
> shoe-horned cbus into ISA bus, and it is a poor fit. Let's learn from
> that lesson.
ACPI isn't a bus like c-bus though. ACPI enumerates ISA devices.
A closer analogy would be to create a pnpbios0 bus for ISA devices
enumerated by the PNP BIOS.
--
John Baldwin <jhb at FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve!" - http://www.FreeBSD.org/
More information about the p4-projects
mailing list