From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeremy Fitzhardinge Subject: Re: RFC: multiple address spaces for one process Date: Sat, 30 Jun 2007 09:30:42 -0700 Message-ID: <46868532.7030208@xensource.com> References: <87myynt1m6.wl%peter@chubb.wattle.id.au> <468517E1.4050803@goop.org> <4685D9C9.20504@de.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <4685D9C9.20504@de.ibm.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: virtualization-bounces@lists.linux-foundation.org Errors-To: virtualization-bounces@lists.linux-foundation.org To: carsteno@de.ibm.com Cc: linux-mm@kvack.org, Peter Chubb , virtualization@lists.linux-foundation.org List-Id: virtualization@lists.linuxfoundation.org Carsten Otte wrote: > The big difference here is that LinuxOnLinux does represent guest > virtual addressing in these mm structs where all other kernel based > VMMs do represent guest physical in the user address space. That > somewhat disqualifies LinuxOnLinux to share the commonality. > Whether or not proposed patch makes sense for shaddow page tables is > unknown to me, since we have nested paging on s390. From an interface perspective, I think nested paging and shadow pagetables should be identical; after all, shadow pagetables are just a software implementation of nested pagetables. There seem to be 3 distinct types of VMM pagetable: 1. UML/LoL vmas-as-pagetable/tlb 2. shadow/nested paging 3. direct paging The multiple address space patch definitely makes sense for 1, but 2&3 both implement the alternate address space by directly pointing the CPU's paging hardware at a new pagetable, rather than going via the Linux VM. J