On Wed, 10 Nov 2004 03:07:40 -0500, Sean Davis <erplefoo@gmail.com> wrote:
> On Wed, 10 Nov 2004 18:18:24 +1100, Dmitri Nikulin
> 
> 
> <setagllib@optusnet.com.au> wrote:
> >
> >
> > Sean Davis wrote:
> >
> > > My main question is... is there any advantage to be gained by enabling
> > > the ioapic support in the kernel? It is currently disabled, and the
> > > box runs just fine... but this is also my first machine to even HAVE
> > > an APIC, so I know jack about it.
> > >
> > > Any advice is appreciated.
> > >
> > Where ACPI configures devices for you, the APIC routes IRQs and handles
> > processor information (vital for SMP, optional for UP). Handy since this
> > usually gets the 'optimal' configuration, and means modern standards (no
> > static BIOS fetching for device configs) are complied with (people who
> > know more than me, please correct this, since I know it's a
> > misunderstanding). Linux, Free and NetBSD all do this now if you have
> > the support in the kernel.
> >
> > I'm not sure if it has any real impact on performance, that's something
> > only a low-level developer could tell you. I certainly haven't noticed
> > any difference. But it does make dmesg look cooler :)
> >
> > How old were your previous machines?
> 
> I've got a 1ghz duron (currently offline, because the motherboard
> likes to reboot randomly), a p3 650mhz, a p3 600mhz, an ultra 5 with a
> 333mhz/2MB cpu module, and an Ultra 1 170E.
> 
> My oldest NetBSD machine is an 80386DX-33 ;-)

BTW: I know none of these, other than perhaps the Duron, have an APIC.
It might. I haven't been motivated enough to boot it and find out, but
ISTR it not having an APIC.

-- 
Sean