HEADSUP: New i386 interrupt and SMP code..

[John Baldwin]

On 31-Oct-2003 David O'Brien wrote:
> On Thu, Oct 30, 2003 at 05:34:47PM -0500, John Baldwin wrote:
>> Downsides:
>> - ACPI will no longer work as a module for know.  The reason
>>   for this is that ACPI's APIC enumerator needs to be able
>>   to hook into a SI_SUB_TUNABLES - 1 SYSINIT() due to existing
>>   code that wants to know the available CPUs in the system
>>   very early (specifically, UMA).
> 
> How will one not have to deal with fscked-up AML BIOS code, when one
> wants to use MADT rather than mptable's to use the APIC?
> 
> I.E., the ACPI-AML bits should be seperatable from the ACPI-MADT bits.

It can't.  The MP Table provides PCI interrupt routing information
in addition to enumerating APICs.  Just look at some 'mptable'
output for comparison.  ACPI's PCI interrupt routing information
is not listed in the MADT, but in _PRT objects in the ACPI namespace.

You could use the MADT w/o ACPI, but then you would have to use no
PCI interrupts.

-- 

John Baldwin <jhb at FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/