[PATCH 00/10] mm: Linux VM Infrastructure to support Memory Power Management

[Ankita Garg]

On Wed, Jul 06, 2011 at 01:20:55PM -0700, david at lang.hm wrote:
> On Wed, 6 Jul 2011, Pekka Enberg wrote:
> 
> >Why does the allocator need to know about address boundaries? Why
> >isn't it enough to make the page allocator and reclaim policies favor using
> >memory from lower addresses as aggressively as possible? That'd mean
> >we'd favor the first memory banks and could keep the remaining ones
> >powered off as much as possible.
> >
> >IOW, why do we need to support scenarios such as this:
> >
> >  bank 0     bank 1   bank 2    bank3
> >| online  | offline | online  | offline |
> 
> I believe that there are memory allocations that cannot be moved
> after they are made (think about regions allocated to DMA from
> hardware where the hardware has already been given the address space
> to DMA into)
>

Thats true. These are kernel allocations which are not movable. However,
the ZONE_MOVABLE would enable us to create complete movable zones and
the ones that have the kernel allocations could be flagged as kernelcore
zone.
 
> As a result, you may not be able to take bank 2 offline, so your
> option is to either leave banks 0-2 all online, or support emptying
> bank 1 and taking it offline.
> 

-- 
Regards,
Ankita Garg (ankita at in.ibm.com)
Linux Technology Center
IBM India Systems & Technology Labs,
Bangalore, India