Messages in this thread | | | Date | Wed, 30 May 2012 21:48:59 +0200 | From | Andi Kleen <> | Subject | Re: [PATCH 2/6] mempolicy: Kill all mempolicy sharing |
| |
On Wed, May 30, 2012 at 02:41:22PM -0500, Christoph Lameter wrote: > On Wed, 30 May 2012, kosaki.motohiro@gmail.com wrote: > > > refcount will be decreased even though was not increased whenever alloc_page_vma() > > is called. As you know, mere mbind(MPOL_MF_MOVE) calls alloc_page_vma(). > > Most of these issues are about memory migration and shared memory. If we > exempt shared memory from memory migration (after all that shared memory > has its own distinct memory policies already!) then a lot of these issues > wont arise.
Soft memory offlining needs migration. It's fairly important that this works: on the database systems most memory is in shared memory and they have a lot of memory, so predictive failure analysis and soft offlining helps a lot.
Classic migration is probably not too important here, but they pretty much rely on the same low level mechanism.
-Andi
-- ak@linux.intel.com -- Speaking for myself only.
| |