This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [Xen-devel] virtual frame_table?

To: Keir Fraser <keir.fraser@xxxxxxxxxxxxx>
Subject: Re: [Xen-devel] virtual frame_table?
From: Alex Williamson <alex.williamson@xxxxxx>
Date: Thu, 27 Aug 2009 08:43:41 -0600
Cc: xen-devel <xen-devel@xxxxxxxxxxxxxxxxxxx>
Delivery-date: Thu, 27 Aug 2009 07:44:07 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
In-reply-to: <C6BBE6A8.1321B%keir.fraser@xxxxxxxxxxxxx>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Organization: HP OSLO R&D
References: <C6BBE6A8.1321B%keir.fraser@xxxxxxxxxxxxx>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
On Thu, 2009-08-27 at 07:30 +0100, Keir Fraser wrote:
> On 26/08/2009 22:20, "Alex Williamson" <alex.williamson@xxxxxx> wrote:
> > On ia64 we had a virtual frame_table to allow for better handling of
> > sparse memory configs.  Has x86 considered anything similar?  I see that
> > init_frametable() is currently allocating 32bytes per page for
> > everything up to max_page.  This can consume a pretty significant chunk
> > of memory if nodes are discontiguous.  Anyone motivated to fix it?
> Easily fixed if there's a reason to. Not really many x86 systems have
> discontiguous memory.

Right, but commodity systems are definitely getting bigger quickly, and
it's only a matter of time.  I'll toss in my vote that the frame_table
should be optimized sooner rather than later.  I might even be able to
test it on a system that's wasting more memory than we'd like there.


Xen-devel mailing list

<Prev in Thread] Current Thread [Next in Thread>