From mboxrd@z Thu Jan 1 00:00:00 1970 From: Konrad Rzeszutek Wilk Subject: Re: (XEN) traps.c:2430:d0 Domain attempted WRMSR 00000000c0010000 from 0x0000000000430076 to 0x0000000000030076. Date: Mon, 15 Aug 2011 11:20:05 -0400 Message-ID: <20110815152005.GA22745@dumpdata.com> References: <4E47ACBC.1040504@it-infrastrukturen.org> <4E48E9940200007800051380@nat28.tlf.novell.com> <20110815123002.GA21495@dumpdata.com> <4E4930F7020000780005147F@nat28.tlf.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <4E4930F7020000780005147F@nat28.tlf.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Jan Beulich Cc: Xen-devel@lists.xensource.com, Mark Schneider , Keir Fraser List-Id: xen-devel@lists.xenproject.org On Mon, Aug 15, 2011 at 01:45:11PM +0100, Jan Beulich wrote: > >>> On 15.08.11 at 14:30, Konrad Rzeszutek Wilk wrote: > > On Mon, Aug 15, 2011 at 08:40:36AM +0100, Jan Beulich wrote: > >> >>> On 14.08.11 at 15:48, Keir Fraser wrote: > >> > You can specify guest_loglvl=error to turn off warning messages, but that > >> > would turn off other messages that might actually be interesting. > >> > Alternatively you'd have to disable the message in source code and rebuild > >> > the hypervisor. Possibly we should reduce this message to debug loglevel > >> > rather than warning logelevel: I agree the message is very much not a > >> > rarity. > >> > >> I tend to disagree - the kernel shouldn't really attempt these writes. > >> One of the things addressing of which is more cumbersome in pv-ops > >> (which I assume is in use here) than "classic" kernels. > > > > It might be good to find out what those calls are and see about actually > > fixing it. > > These are performance monitoring MSRs, and as we seem to agree the > whole perf subsystem is currently not usable with Xen. I would like to see it getting usuable with Xen. Not sure if that is just a "wish" vs a reality. The folks that published the (http://people.cs.vt.edu/~rnikola/files/vee2011.pdf) http://lists.xensource.com/archives/html/xen-devel/2011-05/msg01233.html seemed to have done some work.