Re: [RFC/PATCH 2.6.32] Simple Firmware Interface (SFI): initial support

[Len Brown <lenb@kernel.org>]

On Mon, 22 Jun 2009, Pavel Machek wrote:

> On Tue 2009-06-23 14:41:28, Len Brown wrote:

> > Please let me know if your questions are not thoroughly answered here:
> > http://simplefirmware.org/faq
> 
> It really tells us nothing. I don't think flash got so expensive that
> this is justified. ACPI can already do the job, right? and operating
> systems already have to support ACPI. So what are the reasons to
> reinvent the wheel?

The price of flash, and the amount consumed, is not relevent
to the decision whether a platform should support SFI or ACPI.

The Moorestown platform doesn't use ACPI because its chip-set
fundamentally does not support it.  Not only is the required
register set missing, *all* IO accesses are missing, and there is
no SMM support present to emuate it.

Yes, the ACPI specification could have been edited to replace
every "must" with "could", "shall" with "may", and "required" with 
"optional" resulting in "ACPI compliance" for your toaster.
But doing so would have been a dis-service to the
platforms supporting ACPI, and would have made the
already hard job of supporting ACPI from the OS significantly harder.

thanks,
-Len Brown, Intel Open Source Technology Center