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-ia64-devel] [PATCH 0/16] make xenheap configurable unpinning xe

To: Alex Williamson <alex.williamson@xxxxxx>
Subject: Re: [Xen-ia64-devel] [PATCH 0/16] make xenheap configurable unpinning xenheap
From: Isaku Yamahata <yamahata@xxxxxxxxxxxxx>
Date: Wed, 9 Jan 2008 20:56:15 +0900
Cc: xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
Delivery-date: Wed, 09 Jan 2008 03:56:34 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
In-reply-to: <1199838353.7947.13.camel@lappy>
List-help: <mailto:xen-ia64-devel-request@lists.xensource.com?subject=help>
List-id: Discussion of the ia64 port of Xen <xen-ia64-devel.lists.xensource.com>
List-post: <mailto:xen-ia64-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-ia64-devel>, <mailto:xen-ia64-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-ia64-devel>, <mailto:xen-ia64-devel-request@lists.xensource.com?subject=unsubscribe>
References: <20080107075233.GA13344%yamahata@xxxxxxxxxxxxx> <1199838353.7947.13.camel@lappy>
Sender: xen-ia64-devel-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Mutt/1.5.6i
On Tue, Jan 08, 2008 at 05:25:53PM -0700, Alex Williamson wrote:
> On Mon, 2008-01-07 at 16:52 +0900, Isaku Yamahata wrote:
> >  
> > Presumably benchmarks should be done to confirm that the performance degrade
> > is acceptable.
> Hi Isaku,
>    I tried this patchset, but I can't get xenheap_megabytes to work.  I
> tried 32, 128, 256 and they all hit:
> BUG_ON(!is_xenheap_usable_memory(md))
> in start_kernel().  Does it work for you? 

Could you try again with the attached patch?
It works with 16, 32, 64, 128, 256 and 512. But dom0 panics with 1024.
(The test box is tiger2 with 2GB memory.)
1GB xenheap requirement seems unlikely (at least for now I think)
so that I leave it unfixed.

> I also ran some kernel build
> tests to get an idea of the performance hit.  A UP PV domain was less
> that 1% over my previous tests and an HVM domains was less that 1/4%
> over previous.  I may be into the noise of my test results, but it does
> seem indicate that there is a small hit.  If we can boot more vCPUs and
> bigger systems though, it may be a reasonable trade-off.  Thanks,

Good news. thank you for testing.

Attachment: 16712_3f8403f8ae2e_boot_load_data_start_kernel.patch
Description: Text Data

Xen-ia64-devel mailing list