From mboxrd@z Thu Jan  1 00:00:00 1970
From: Joerg Roedel <joro@8bytes.org>
Subject: Re: Fwd: [KVM TSC emulation 6/9] Allow adjust_tsc_offset to be in
	host or guest cycles
Date: Wed, 6 Jul 2011 12:00:11 +0200
Message-ID: <20110706100010.GL29299@8bytes.org>
References: <4E008D05.8080705@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: kvm <kvm@vger.kernel.org>
To: Zachary Amsden <zamsden@redhat.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from 8bytes.org ([88.198.83.132]:35160 "EHLO 8bytes.org"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1750977Ab1GFKAM (ORCPT <rfc822;kvm@vger.kernel.org>);
	Wed, 6 Jul 2011 06:00:12 -0400
Content-Disposition: inline
In-Reply-To: <4E008D05.8080705@redhat.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Tue, Jun 21, 2011 at 05:22:29AM -0700, Zachary Amsden wrote:
> -static void svm_adjust_tsc_offset(struct kvm_vcpu *vcpu, s64 adjustment)
> +static void svm_adjust_tsc_offset(struct kvm_vcpu *vcpu, s64 adjustment, bool host)
>  {
>  	struct vcpu_svm *svm = to_svm(vcpu);
>
> +	WARN_ON(adjustment<  0);
> +	if (host)
> +		adjustment = svm_scale_tsc(vcpu, adjustment);
> +

This is not going to work out with tsc-scaling. The tsc-offset is
applied _after_ the hardware-tsc. Just scaling the offset with the same
factor will not give you the hardware tsc-value you expect.

	Joerg