[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Xen-devel] Re: [RFC PATCH 0/4] (Take 2): transcendent memory ("tmem") for Linux
- To: Dan Magenheimer <dan.magenheimer@xxxxxxxxxx>
- From: Anthony Liguori <anthony@xxxxxxxxxxxxx>
- Date: Sun, 12 Jul 2009 08:28:34 -0500
- Cc: npiggin@xxxxxxx, akpm@xxxxxxxx, xen-devel@xxxxxxxxxxxxxxxxxxx, tmem-devel@xxxxxxxxxxxxxx, kurt.hackel@xxxxxxxxxx, Rusty Russell <rusty@xxxxxxxxxxxxxxx>, linux-kernel@xxxxxxxxxxxxxxx, dave.mccracken@xxxxxxxxxx, linux-mm@xxxxxxxxx, chris.mason@xxxxxxxxxx, sunil.mushran@xxxxxxxxxx, Avi Kivity <avi@xxxxxxxxxx>, jeremy@xxxxxxxx, Schwidefsky <schwidefsky@xxxxxxxxxx>, Marcelo Tosatti <mtosatti@xxxxxxxxxx>, alan@xxxxxxxxxxxxxxxxxxx, Balbir Singh <balbir@xxxxxxxxxxxxxxxxxx>
- Delivery-date: Sun, 12 Jul 2009 06:29:06 -0700
- List-id: Xen developer discussion <xen-devel.lists.xensource.com>
Dan Magenheimer wrote:
Oops, sorry, I guess that was a bit inflammatory. What I meant to
say is that inferring resource utilization efficiency is a very
hard problem and VMware (and I'm sure IBM too) has done a fine job
with it; CMM2 explicitly provides some very useful information from
within the OS to the hypervisor so that it doesn't have to infer
that information; but tmem is trying to go a step further by making
the cooperation between the OS and hypervisor more explicit
and directly beneficial to the OS.
KVM definitely falls into the camp of trying to minimize modification to
If there was one change to tmem that would make it more
me it would be changing the way pools are "allocated". Instead of
getting an opaque handle from the hypervisor, I would force
the guest to
allocate it's own memory and to tell the hypervisor that it's a tmem
An interesting idea but one of the nice advantages of tmem being
completely external to the OS is that the tmem pool may be much
larger than the total memory available to the OS. As an extreme
example, assume you have one 1GB guest on a physical machine that
has 64GB physical RAM. The guest now has 1GB of directly-addressable
memory and 63GB of indirectly-addressable memory through tmem.
That 63GB requires no page structs or other data structures in the
guest. And in the current (external) implementation, the size
of each pool is constantly changing, sometimes dramatically so
the guest would have to be prepared to handle this. I also wonder
if this would make shared-tmem-pools more difficult.
I can see how it might be useful for KVM though. Once the
core API and all the hooks are in place, a KVM implementation of
tmem could attempt something like this.
It's the core API that is really the issue. The semantics of tmem
(external memory pool with copy interface) is really what is problematic.
The basic concept, notifying the VMM about memory that can be recreated
by the guest to avoid the VMM having to swap before reclaim, is great
and I'd love to see Linux support it in some way.
The big advantage of keeping the tmem pool part of the normal set of
guest memory is that you don't introduce new challenges with
respect to memory accounting. Whether or not tmem is directly
accessible from the guest, it is another memory resource. I'm
certain that you'll want to do accounting of how much tmem is being
consumed by each guest
Yes, the Xen implementation of tmem does accounting on a per-pool
and a per-guest basis and exposes the data via a privileged
"tmem control" hypercall.
I was talking about accounting within the guest. It's not just a matter
of accounting within the mm, it's also about accounting in userspace. A
lot of software out there depends on getting detailed statistics from
Linux about how much memory is in use in order to determine things like
memory pressure. If you introduce a new class of memory, you need a new
class of statistics to expose to userspace and all those tools need
Xen-devel mailing list