UP IO-APIC

UP IO-APIC

[Ricky Beam]

Would someone kindly remove that from the configuration possibilities?  It
doesn't work -- and hasn't worked for, what, a year.

--Ricky

Re: UP IO-APIC

[Miquel van Smoorenburg]

In article <Pine.GSO.4.33.0209231419230.11624-100000@sweetums.bluetronic.net>,
Ricky Beam  <jfbeam@bluetronic.net> wrote:
>Would someone kindly remove that from the configuration possibilities?  It
>doesn't work -- and hasn't worked for, what, a year.

What kernel? In 2.4.20-pre<whatever> it works just fine. If it
doesn't work in 2.5 leave it there as an incentive for someone
to fix it for 2.6

Mike.
-- 
Computers are useless, they only give answers. --Pablo Picasso

Re: UP IO-APIC

[Zwane Mwaikambo]

On Mon, 23 Sep 2002, Ricky Beam wrote:

> Would someone kindly remove that from the configuration possibilities?  It
> doesn't work -- and hasn't worked for, what, a year.

Worked for me on 2.4.19-pre and 2.5 (haven't tried recent), and still 
thats a bad reason to remove it.

	Zwane

-- 
function.linuxpower.ca

Re: UP IO-APIC

[Ricky Beam]

On Tue, 24 Sep 2002, Zwane Mwaikambo wrote:
>> Would someone kindly remove that from the configuration possibilities?  It
>> doesn't work -- and hasn't worked for, what, a year.
>
>Worked for me on 2.4.19-pre and 2.5 (haven't tried recent), and still
>thats a bad reason to remove it.

It works in 2.4, but I've never seen it work in 2.5 -- but I've not compiled
every 2.5.X.  Neither the local APIC or IO APIC work in non-SMP configurations
due to dependencies on things in mpparse.c (read: SMP functions.)  The local
APIC makes perfect sense albeit rare.  Single processor IO APICs are very
rare and are usually MP systems with only one processor.

APIC support in 2.5 is very closely tied to SMP. (and technically, ACPI.)

--Ricky

Re: UP IO-APIC

[Miquel van Smoorenburg]

In article <Pine.GSO.4.33.0209241119500.11624-100000@sweetums.bluetronic.net>,
Ricky Beam  <jfbeam@bluetronic.net> wrote:
>The local
>APIC makes perfect sense albeit rare.  Single processor IO APICs are very
>rare and are usually MP systems with only one processor.

I think most AMD Athlon boards have an IO APIC

Mike.
-- 
Computers are useless, they only give answers. --Pablo Picasso

Re: UP IO-APIC

[Zwane Mwaikambo]

On Tue, 24 Sep 2002, Ricky Beam wrote:

> It works in 2.4, but I've never seen it work in 2.5 -- but I've not compiled
> every 2.5.X.  Neither the local APIC or IO APIC work in non-SMP configurations

I've been using it since inception with local APIC, UP-IOAPIC broke a 
couple of times only.

> due to dependencies on things in mpparse.c (read: SMP functions.)  The local
> APIC makes perfect sense albeit rare.  Single processor IO APICs are very
> rare and are usually MP systems with only one processor.

... Or UP machines with chipsets with IOAPICs (Microsoft recommends 
hardware manufaturers to use one).

> APIC support in 2.5 is very closely tied to SMP. (and technically, ACPI.)

Not the case for local APIC, i'll try building .38 with UP IOAPIC.

	Zwane

-- 
function.linuxpower.ca

Re: UP IO-APIC

[Bob_Tracy]

Zwane Mwaikambo wrote:
> On Tue, 24 Sep 2002, Ricky Beam wrote:
> 
> > It works in 2.4, but I've never seen it work in 2.5 -- but I've not compiled
> > every 2.5.X.  Neither the local APIC or IO APIC work in non-SMP configurations
> 
> I've been using it since inception with local APIC, UP-IOAPIC broke a 
> couple of times only.

My Dell notebook has an APIC, but it has gotten disabled automatically
at boot time ever since someone labeled my BIOS as broken :-).  FWIW,
I don't disagree with the "broken" label.

-- 
-----------------------------------------------------------------------
Bob Tracy                   WTO + WIPO = DMCA? http://www.anti-dmca.org
rct@frus.com
-----------------------------------------------------------------------

Re: UP IO-APIC

[Bill Davidsen]

On Tue, 24 Sep 2002, Zwane Mwaikambo wrote:

> 
> ... Or UP machines with chipsets with IOAPICs (Microsoft recommends 
> hardware manufaturers to use one).
> 
> > APIC support in 2.5 is very closely tied to SMP. (and technically, ACPI.)
> 
> Not the case for local APIC, i'll try building .38 with UP IOAPIC.

I didn't get a compile with it on, finally got a compile and boot of
2.5.38-mm2 with most of the patches here in the last few days. It doesn't
see my SCSI controller, sound doesn't work, but it booted ;-)

-- 
bill davidsen <davidsen@tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

Re: UP IO-APIC

[Benjamin LaHaise]

On Tue, Sep 24, 2002 at 06:01:10PM +0000, Miquel van Smoorenburg wrote:
> In article <Pine.GSO.4.33.0209241119500.11624-100000@sweetums.bluetronic.net>,
> Ricky Beam  <jfbeam@bluetronic.net> wrote:
> >The local
> >APIC makes perfect sense albeit rare.  Single processor IO APICs are very
> >rare and are usually MP systems with only one processor.
> 
> I think most AMD Athlon boards have an IO APIC

I'd love to have it enabled in a distro kernel, but as Arjan pointed out, it 
currently breaks some laptops if enabled.  What we need is someone to weed 
things out such that io apic setup gets done after command line parsing, but 
that is a bit tricky...

		-ben

Re: UP IO-APIC

[Lars Marowsky-Bree]

On 2002-09-24T19:19:34,
   Benjamin LaHaise <bcrl@redhat.com> said:

> > >APIC makes perfect sense albeit rare.  Single processor IO APICs are very
> > >rare and are usually MP systems with only one processor.
> > I think most AMD Athlon boards have an IO APIC
> I'd love to have it enabled in a distro kernel, but as Arjan pointed out, it 
> currently breaks some laptops if enabled.

Well, _not_ enabling IO-APIC on UP breaks my Athlon / KT333 at home; random
freezes are the result, so I prefer to enable it...

So whichever default is chosen, someone is burned. I hate hardware.


Sincerely,
    Lars Marowsky-Brée <lmb@suse.de>

-- 
Principal Squirrel
Research and Development, SuSE Linux AG
 
``Immortality is an adequate definition of high availability for me.''
	--- Gregory F. Pfister