From mboxrd@z Thu Jan  1 00:00:00 1970
From: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Subject: Re: [PATCH v2 06/14] x86/time.c: Scale host TSC in
	pvclock properly
Date: Mon, 7 Dec 2015 13:14:54 -0500
Message-ID: <5665CC9E.4040302@oracle.com>
References: <1449435529-12989-1-git-send-email-haozhong.zhang@intel.com>
	<1449435529-12989-7-git-send-email-haozhong.zhang@intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; Format="flowed"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xen.org>
In-Reply-To: <1449435529-12989-7-git-send-email-haozhong.zhang@intel.com>
List-Unsubscribe: <http://lists.xen.org/cgi-bin/mailman/options/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xen.org>
List-Help: <mailto:xen-devel-request@lists.xen.org?subject=help>
List-Subscribe: <http://lists.xen.org/cgi-bin/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=subscribe>
Sender: xen-devel-bounces@lists.xen.org
Errors-To: xen-devel-bounces@lists.xen.org
To: Haozhong Zhang <haozhong.zhang@intel.com>, xen-devel@lists.xen.org
Cc: Kevin Tian <kevin.tian@intel.com>, Keir Fraser <keir@xen.org>, Jan Beulich <jbeulich@suse.com>, Jun Nakajima <jun.nakajima@intel.com>, Andrew Cooper <andrew.cooper3@citrix.com>, Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>, Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>, Joao Martins <joao.m.martins@oracle.com>
List-Id: xen-devel@lists.xenproject.org

On 12/06/2015 03:58 PM, Haozhong Zhang wrote:
> This patch makes the pvclock return the scaled host TSC and
> corresponding scaling parameters to HVM domains if guest TSC is not
> emulated and TSC scaling is enabled.
>
> Signed-off-by: Haozhong Zhang <haozhong.zhang@intel.com>

+Joao who has been staring at this code.

Joao, can you run this series through your test with non-native 
frequency? 
(http://lists.xenproject.org/archives/html/xen-devel/2015-12/msg00683.html 
provides an interface to set it in config file).


> ---
>   xen/arch/x86/time.c | 16 ++++++++++++----
>   1 file changed, 12 insertions(+), 4 deletions(-)
>
> diff --git a/xen/arch/x86/time.c b/xen/arch/x86/time.c
> index 95df4f1..732d1e9 100644
> --- a/xen/arch/x86/time.c
> +++ b/xen/arch/x86/time.c
> @@ -815,10 +815,18 @@ static void __update_vcpu_system_time(struct vcpu *v, int force)
>       }
>       else
>       {
> -        tsc_stamp = t->local_tsc_stamp;
> -
> -        _u.tsc_to_system_mul = t->tsc_scale.mul_frac;
> -        _u.tsc_shift         = (s8)t->tsc_scale.shift;
> +        if ( is_hvm_domain(d) && cpu_has_tsc_ratio )
> +        {
> +            tsc_stamp            = hvm_funcs.scale_tsc(v, t->local_tsc_stamp);
> +            _u.tsc_to_system_mul = d->arch.vtsc_to_ns.mul_frac;
> +            _u.tsc_shift         = d->arch.vtsc_to_ns.shift;

I am not sure this is correct (which is why I asked Joao to look at this 
and test). The scaler below is calculated as result of TSC calibration 
across physical CPUs and what you use above (vtsc_to_ns) is an 
uncalibrated value.

-boris

> +        }
> +        else
> +        {
> +            tsc_stamp            = t->local_tsc_stamp;
> +            _u.tsc_to_system_mul = t->tsc_scale.mul_frac;
> +            _u.tsc_shift         = (s8)t->tsc_scale.shift;
> +        }
>       }
>   
>       _u.tsc_timestamp = tsc_stamp;