Date: Wed, 6 Dec 2017 09:55:43 -0800
From: Andi Kleen <ak@linux.intel.com>
To: Jan Dakinevich <jan.dakinevich@virtuozzo.com>
Cc: linux-kernel@vger.kernel.org, "Denis V . Lunev" <den@virtuozzo.com>,
        Roman Kagan <rkagan@virtuozzo.com>,
        Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        Alexander Shishkin <alexander.shishkin@linux.intel.com>,
        Jiri Olsa <jolsa@redhat.com>, Namhyung Kim <namhyung@kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>, "H. Peter Anvin" <hpa@zytor.com>,
        x86@kernel.org, Paolo Bonzini <pbonzini@redhat.com>,
        Radim =?utf-8?B?S3LEjW3DocWZ?= <rkrcmar@redhat.com>,
        Kan Liang <kan.liang@intel.com>, Stephane Eranian <eranian@google.com>,
        Colin King <colin.king@canonical.com>,
        Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
        Jin Yao <yao.jin@linux.intel.com>, kvm@vger.kernel.org
Subject: Re: [PATCH RFC 2/2] KVM: x86/vPMU: ignore access to LBR-related MSRs
Message-ID: <20171206175543.GM3070@tassilo.jf.intel.com>
References: <1512560585-27263-1-git-send-email-jan.dakinevich@virtuozzo.com>
 <1512560585-27263-3-git-send-email-jan.dakinevich@virtuozzo.com>
 <20171206155728.GL3070@tassilo.jf.intel.com>
 <20171206200207.11094037@virtuozzo.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20171206200207.11094037@virtuozzo.com>
User-Agent: Mutt/1.9.1 (2017-09-22)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 944
Lines: 25

On Wed, Dec 06, 2017 at 08:02:07PM +0300, Jan Dakinevich wrote:
> On Wed, 6 Dec 2017 07:57:28 -0800
> Andi Kleen <ak@linux.intel.com> wrote:
> 
> > 
> > If you do all this it's only a small step to fully enable LBRs for
> > guests. 
> 
> It is quite simple in a case where guest LBR-related MSRs matches host
> ones. They could be handled by MSR load/store areas, I suppose.

There is already a LBR control to enable/disable I believe.
You don't want to save/restore all MSRs on every entry/exit 
because that would be slow. The normal Linux context switch can do it.
 
> 
> In other cases, it could be expected the different amount of these MSRs
> and different theirs base values (e.g. Nehalem vs Core). Guest MSRs
> could be both subset and superset of host MSRs, so additional efforts
> to support this would be required.

In this case ignoring would be sufficient I suppose. But for the case
when everything matches it should work.

-Andi