Re: [PATCH 1/2] mm: Export vm_committed_as

[Andrew Morton <akpm@linux-foundation.org>]

On Mon, 8 Oct 2012 06:35:39 -0700
Greg KH <gregkh@linuxfoundation.org> wrote:

> On Mon, Oct 08, 2012 at 03:35:50AM +0000, KY Srinivasan wrote:
> > 
> > 
> > > -----Original Message-----
> > > From: Greg KH [mailto:gregkh@linuxfoundation.org]
> > > Sent: Sunday, October 07, 2012 8:44 PM
> > > To: KY Srinivasan
> > > Cc: linux-kernel@vger.kernel.org; devel@linuxdriverproject.org; olaf@aepfle.de;
> > > apw@canonical.com; akpm@linux-foundation.org; andi@firstfloor.org
> > > Subject: Re: [PATCH 1/2] mm: Export vm_committed_as
> > > 
> > > On Sun, Oct 07, 2012 at 04:59:45PM -0700, K. Y. Srinivasan wrote:
> > > > The policy engine on the host expects the guest to report the
> > > > committed_as. Since this variable is not exported,
> > > > export this symbol.
> > > 
> > > Why are these symbols not needed by either Xen or KVM or vmware, which
> > > I think all support the same thing, right?
> > 
> > The basic balloon driver does not need this symbol since the basic balloon driver
> > is not automatically driven by the host. On the Windows host we have a policy engine that
> > drives the balloon driver based on both guest level memory pressure that the guest
> > reports as well as other system level metrics the host maintains. We need this symbol to
> > drive the policy engine on the host.
> 
> Ok, but you're going to have to get the -mm developers to agree that
> this is ok before I can accept it.

Well I guess it won't kill us.

I do wonder what relevance vm_committed_as has to processes which are
running within a memcg container?