Re: RFC: NUMA support

[Christoph Egger <Christoph_Egger%gmx.de@localhost>]

On Monday 10 November 2008 16:36:15 Manuel Bouyer wrote:
> On Mon, Nov 10, 2008 at 05:11:37PM +0200, Christoph Egger wrote:
> > Hi!
> >
> > I started to work on NUMA support. First step is to set up the
> > topology.
> >
> > It does this by scanning the ACPI SRAT table.
> > If no ACPI SRAT table is present or if you boot w/ ACPI disabled,
> > a one node NUMA system is faked.
> >
> > The boot code also utilizes the ACPI MADT table to get more
> > information so far possible.
> >
> > I showed rmind@ my patch so far. I share his opinion, that
> > it needs some more thoughts on the MI API side.
>
> FYI, at work we're working on a project which targets an architecture with
> up to 4k CPU core, heavily NUMA (the L2 caches can't cache the whole memory
> range). so I'm all for a good MI API on this :)

The affinity information are very important on such large machines.
The ACPI SLIT table exists on nearly no machines. I need to have a look
at how Solaris is doing the measurement.

bouyer: Is your machine a SGI machine ?

> > [...]
> > A note, this patch breaks compiling Xen2 kernels. But since Xen2 is going
> > to be removed in -current, I don't care.
>
> But it's not removed yet. it's be better to have it build for now
> (especially as I suspect it shouldn't be hard to make it compile).

Please, let's discuss this further on port-xen with the
subject "Re: Xen2 removal: now or later".

Christoph