From mboxrd@z Thu Jan  1 00:00:00 1970
From: Roman Kagan <rkagan@virtuozzo.com>
Subject: Re: [PATCH v2 1/2] x86:kvm:hyperv: add KVM_CAP_HYPERV_SYNIC2
Date: Thu, 22 Jun 2017 08:14:14 +0300
Message-ID: <20170622051411.GA15650@rkaganip.lan>
References: <20170616173740.23272-1-rkagan@virtuozzo.com>
 <20170616173740.23272-2-rkagan@virtuozzo.com>
 <20170621200915.GD27032@potion>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Cc: kvm@vger.kernel.org, Paolo Bonzini <pbonzini@redhat.com>,
        Evgeny Yakovlev <eyakovlev@virtuozzo.com>,
        "Denis V . Lunev" <den@openvz.org>,
        Eduardo Habkost <ehabkost@redhat.com>,
        Igor Mammedov <imammedo@redhat.com>
To: Radim =?utf-8?B?S3LEjW3DocWZ?= <rkrcmar@redhat.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mail-eopbgr50128.outbound.protection.outlook.com ([40.107.5.128]:4032
        "EHLO EUR03-VE1-obe.outbound.protection.outlook.com"
        rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP
        id S1751064AbdFVFOh (ORCPT <rfc822;kvm@vger.kernel.org>);
        Thu, 22 Jun 2017 01:14:37 -0400
Content-Disposition: inline
In-Reply-To: <20170621200915.GD27032@potion>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Wed, Jun 21, 2017 at 10:09:15PM +0200, Radim Krčmář wrote:
> 2017-06-16 20:37+0300, Roman Kagan:
> > There is a flaw in the Hyper-V SynIC implementation in KVM: when message
> > page or event flags page is enabled by setting the corresponding msr,
> > KVM zeroes it out.  This is problematic because on migration the
> > corresponding MSRs are loaded on the destination, so the content of
> > those pages is lost.
> > 
> > This went unnoticed so far because the only user of those pages was
> > in-KVM hyperv synic timers, which could continue working despite that
> > zeroing.
> > 
> > Newer QEMU uses those pages for Hyper-V VMBus implementation, and
> > zeroing them breaks the migration.
> > 
> > Besides, in newer QEMU the content of those pages is fully managed by
> > QEMU, so zeroing them is undesirable even when writing the MSRs from the
> > guest side.
> > 
> > To support this new scheme, introduce a new capability,
> > KVM_CAP_HYPERV_SYNIC2, which, when enabled, makes sure that the synic
> > pages aren't zeroed out in KVM.
> > 
> > Signed-off-by: Roman Kagan <rkagan@virtuozzo.com>
> > ---
> 
> I have changed the subject tags to more common "kvm: x86: hyperv:" and
> added minimal documentation:
> 
> diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
> index 4029943887a3..3f4b1d5f0dce 100644
> --- a/Documentation/virtual/kvm/api.txt
> +++ b/Documentation/virtual/kvm/api.txt
> @@ -4157,3 +4157,10 @@ Currently the following bits are defined for the device_irq_level bitmap:
>  Future versions of kvm may implement additional events. These will get
>  indicated by returning a higher number from KVM_CHECK_EXTENSION and will be
>  listed above.
> +
> +8.10 KVM_CAP_HYPERV_SYNIC2
> +
> +Architectures: x86
> +
> +The only difference from KVM_CAP_HYPERV_SYNIC is that KVM does not clear SynIC
> +pages when the guest enables them.
> 
> Please let me know if you'd like to improve it.

Yes that's perfectly fine, thanks.  Sorry, I should have done it
myself...

Thanks,
Roman.