From mboxrd@z Thu Jan  1 00:00:00 1970
From: Avi Kivity <avi@redhat.com>
Subject: Re: [RFC] para virt interface of perf to support kvm guest os statistics
 collection in guest os
Date: Wed, 09 Jun 2010 11:59:48 +0300
Message-ID: <4C0F5804.9080406@redhat.com>
References: <1276054214.2096.383.camel@ymzhang.sh.intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: LKML <linux-kernel@vger.kernel.org>, kvm@vger.kernel.org,
	Ingo Molnar <mingo@elte.hu>,
	Fr??d??ric Weisbecker <fweisbec@gmail.com>,
	Arnaldo Carvalho de Melo <acme@redhat.com>,
	Cyrill Gorcunov <gorcunov@gmail.com>,
	Lin Ming <ming.m.lin@intel.com>,
	Sheng Yang <sheng@linux.intel.com>,
	Marcelo Tosatti <mtosatti@redhat.com>,
	oerg Roedel <joro@8bytes.org>,
	Jes Sorensen <Jes.Sorensen@redhat.com>,
	Gleb Natapov <gleb@redhat.com>,
	Zachary Amsden <zamsden@redhat.com>, zhiteng.huang@intel.com,
	tim.c.chen@intel.com
To: "Zhang, Yanmin" <yanmin_zhang@linux.intel.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mx1.redhat.com ([209.132.183.28]:46272 "EHLO mx1.redhat.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1752994Ab0FIJAH (ORCPT <rfc822;kvm@vger.kernel.org>);
	Wed, 9 Jun 2010 05:00:07 -0400
In-Reply-To: <1276054214.2096.383.camel@ymzhang.sh.intel.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On 06/09/2010 06:30 AM, Zhang, Yanmin wrote:
> From: Zhang, Yanmin<yanmin_zhang@linux.intel.com>
>
> Based on Ingo's idea, I implement a para virt interface for perf to support
> statistics collection in guest os. That means we could run tool perf in guest
> os directly.
>
> Great thanks to Peter Zijlstra. He is really the architect and gave me architecture
> design suggestions. I also want to thank Yangsheng and LinMing for their generous
> help.
>
> The design is:
>
> 1) Add a kvm_pmu whose callbacks mostly just calls hypercall to vmexit to host kernel;
> 2) Create a host perf_event per guest perf_event;
> 3) Host kernel syncs perf_event count/overflows data changes to guest perf_event
> when processing perf_event overflows after NMI arrives. Host kernel inject NMI to guest
> kernel if a guest event overflows.
> 4) Guest kernel goes through all enabled event on current cpu and output data when they
> overflows.
> 5) No change in user space.
>    

Other issues:

- save/restore support for live migration
- some way to limit the number of open handles (comes automatically with 
the table approach I suggested earlier)

-- 
error compiling committee.c: too many arguments to function