WARNING - OLD ARCHIVES

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/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-ia64-devel

Re: [Xen-ia64-devel] psr.sp/is/di/si virtualization

To: Tristan Gingold <tgingold@xxxxxxx>
Subject: Re: [Xen-ia64-devel] psr.sp/is/di/si virtualization
From: Isaku Yamahata <yamahata@xxxxxxxxxxxxx>
Date: Wed, 9 May 2007 11:51:03 +0900
Cc: Xen-ia64-devel <xen-ia64-devel@xxxxxxxxxxxxxxxxxxx>
Delivery-date: Tue, 08 May 2007 19:49:51 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
In-reply-to: <20070508133838.GA16105@saphi>
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: <20070508133838.GA16105@saphi>
Sender: xen-ia64-devel-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Mutt/1.4.2.1i
On Tue, May 08, 2007 at 03:38:38PM +0200, Tristan Gingold wrote:

> these psr bits are not fully virtualized and the current usage is not
> clear (at least to me):
> 
> sp (secure performance monitor): I don't know how it is currently used by
>   xenoprofile.  Isaku have you any tips on it ?

The current xenoprofile model is that xen vmm controls pmc/pmd.
So xenoprof needs to prevent the guest os changing psr.up bit
(user pserfomance monitor enable) by setting pser.sp.

When considering pmc/pmd virtualization, the xenoprof requirement
should be revised. How the virtualization coexist with
xenoprof should be considered.


> di (disable instruction set transition): should it be hard-coded to 1 or
>  should it be settable by the kernel ?
> 
> si (secure interval timer): should be forced to 0?

Vanilla Linux/ia64 sets psr.si = 0 so that application can read it
without kernel intervention.
But I don't know how many application expect this fact.
Probably only benchmarking apps?


> is (instruction set): same use as di.
> 
> mc (machine check): should be forced to 1?

Or dom0 desires to change?

-- 
yamahata

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel

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