RE: [patch 2/2] add x86-64 support for memory hot-add

[keith]

On Mon, 2006-01-09 at 11:55 -0800, Tolentino, Matthew E wrote:
> Andi Kleen <mailto:ak@xxxxxxx> wrote:
> > On Monday 09 January 2006 16:21, Matt Tolentino wrote:
> >> Add x86-64 specific memory hot-add functions, Kconfig options,
> >> and runtime kernel page table update functions to make
> >> hot-add usable on x86-64 machines.  Also, fixup the nefarious
> >> conditional locking and exports pointed out by Andi.
> > 
> > I'm trying to stabilize my tree for the 2.6.16 submission right now
> > and this one comes a bit too late and is a bit too involved
> > to slip through - sorry. I will consider it after Linus
> > has merged the whole batch of changes for 2.6.16 - so hopefully
> > in 2.6.17.
> > 
> >> +/*
> >> + * Memory hotplug specific functions
> >> + * These are only for non-NUMA machines right now.
> > 
> > How much work would it be to allow it for NUMA kernels too?

Not too much.  I have a start of this code.  Just saving off the SRAT
locality information and using it during the add-event to decide which
node it goes to. But I went to test this weekend on a multi-node system
and the underlying __add_pages refused to add the pages.  The underlying
sparsemem can do this (It works for PPC).  I am still collecting the
debug info needed.

When I get the numa system sorted I will post patches. 

-- 
keith <kmannth@xxxxxxxxxx>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/