From mboxrd@z Thu Jan  1 00:00:00 1970
From: Keir Fraser <keir.fraser@eu.citrix.com>
Subject: Re: Dump VMCS on Intel CPU
Date: Sat, 06 Jun 2009 07:23:32 +0100
Message-ID: <C64FC9F4.CF97%keir.fraser@eu.citrix.com>
References: <4A29DD42.5080200@oracle.com>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xensource.com>
In-Reply-To: <4A29DD42.5080200@oracle.com>
List-Unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xensource.com>
List-Help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-Subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
Sender: xen-devel-bounces@lists.xensource.com
Errors-To: xen-devel-bounces@lists.xensource.com
To: "mukesh.rathor@oracle.com" <mukesh.rathor@oracle.com>, "xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>
List-Id: xen-devel@lists.xenproject.org

On 06/06/2009 04:06, "Mukesh Rathor" <mukesh.rathor@oracle.com> wrote:

> On AMD, it was simple by calling svm_dump_vmcb() by passing vmcb from
> each vcpu struct. Looking at vmcs.c, I see quite a bit of mumbo-jumbo in
> terms of setting context, even IPI other CPUs, etc. That doesn't work
> for me. So the question is, is there any reason followig won't work?
> 
>    for each hvm vcpu {
>        __vmptrld(virt_to_maddr(vp->arch.hvm_vmx.vmcs));
>        vmcs_dump_vcpu();
>    }
>    if ( is_hvm_vcpu(current) )
>        __vmptrld(virt_to_maddr(current->arch.hvm_vmx.vmcs));
> 
> 
> Seems to work ok, but I want to make sure.

Intel say you should VMCLEAR on the last CPU to VMPTRLD the VMCS. For us
that means if v->arch.hvm_vmx.active_cpu != smp_processor_id() nor -1.
However VMPTRLD is not described as failing if this is not carried out. So,
I guess you may get away with it. If you want to be super safe you might
VMCLEAR on CPUs as they rendezvous with kdb.

 -- Keir