[PATCH] PPC64 NUMA memory fixup

mike kravetz kravetz at us.ibm.com
Fri Mar 11 10:42:47 EST 2005


On Thu, Mar 10, 2005 at 02:36:13AM -0800, Andrew Morton wrote:
> Paul Mackerras <paulus at samba.org> wrote:
> >
> > When I booted my new 720 on a kernel configured for NUMA, I received
> >  the following during bootup:
> > 
> >  WARNING: Unexpected node layout: region start 44000000 length 2000000
> >  NUMA is disabled
> > 
> >  This is due to memory 'holes' within nodes.  If such holes are
> >  encountered, then NUMA is disabled.  The following patch adds support
> >  for such configurations.  My 720 now boots with the following message:
> 
> This patch causes the non-numa G5 to oops very early in boot in
> smp_call_function().
> 
<Andrew's config file removed>

I can't recreate this on my system here even if I start with Andrew's
config file.  In addition, I don't have access to a PMAC to even try
and figure out the flow at boot time.  My guess is that this is related
to the extra scan of memory sections (via of_find_node_by_type()) in
do_init_bootmem.  Is there something inherently different between making
these calls on a LPAR as opposed to PMAC?  I'm going to start looking
for a PMAC so I can get more info.  Any other suggestions on how to
track this down are appreciated.

Thanks,
--
Mike



More information about the Linuxppc64-dev mailing list