Ok, here's a new patch that addresses the problems below, just for
reference.  I'll split out the contig.c stuff and send that seperately
in a minute.

On Mon, Sep 22, 2003 at 12:39:40PM -0700, David Mosberger wrote:
>  - in acpi.c, it looks like there is no reason for min_hole_size anymore,
>    yet the patch doesn't clean up this variable

I had just removed that printk at Xavier's request and missed that it
wasn't being used anymore.  Fixed.

>  - MAX_RSVD_REGIONS et al got moved to pgtable.h; why in the world would
>    that be the right place?  Just because "it compiled"??

No, because that was necessary for the contig.c/discontig.c seperation.
Which header file would be better?

>  - why is there both vmem_map and mem_map in the NUMA case?

You mean for the discontig case?  mem_map isn't there.  It only exists
(mm/memory.c) if CONFIG_DISCONTIGMEM is off, otherwise we just use
vmem_map and pgdata_t->node_mem_map for everything.

Right now (and even without this patch) the config options are kind of a
mess.  CONFIG_DISCONTIGMEM doesn't make sense without CONFIG_NUMA and
CONFIG_VIRTUAL_MEMMAP even though they should all be independent.  I'll
try to clean that up too.

Thanks,
Jesse