Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Content-Type: text/plain;
        charset=us-ascii
Mime-Version: 1.0 (1.0)
Subject: Re: [PATCH] x86: vmx: Allow direct access to MSR_IA32_SPEC_CTRL
From:   Andy Lutomirski <luto@amacapital.net>
In-Reply-To: <4DCAF18F-C86A-4CBC-A9CC-CC01BF63313F@oracle.com>
Date:   Sun, 28 Jan 2018 12:40:38 -0800
Cc:     KarimAllah Ahmed <karahmed@amazon.de>, kvm@vger.kernel.org,
        linux-kernel@vger.kernel.org,
        Asit Mallick <asit.k.mallick@intel.com>,
        Arjan Van De Ven <arjan.van.de.ven@intel.com>,
        Dave Hansen <dave.hansen@intel.com>,
        Andi Kleen <ak@linux.intel.com>,
        Andrea Arcangeli <aarcange@redhat.com>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Tim Chen <tim.c.chen@linux.intel.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Dan Williams <dan.j.williams@intel.com>,
        Jun Nakajima <jun.nakajima@intel.com>,
        Paolo Bonzini <pbonzini@redhat.com>,
        David Woodhouse <dwmw@amazon.co.uk>,
        Greg KH <gregkh@linuxfoundation.org>,
        Andy Lutomirski <luto@kernel.org>,
        Ashok Raj <ashok.raj@intel.com>, daniel.kiper@oracle.com
Content-Transfer-Encoding: quoted-printable
Message-Id: <86F5D9C0-7B11-448D-954F-D2EF6CBE80EC@amacapital.net>
References: <1517167750-23485-1-git-send-email-karahmed@amazon.de> <4DCAF18F-C86A-4CBC-A9CC-CC01BF63313F@oracle.com>
To:     Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Jan 28, 2018, at 12:21 PM, Konrad Rzeszutek Wilk <konrad.wilk@oracle.co=
m> wrote:
>=20
>> On January 28, 2018 2:29:10 PM EST, KarimAllah Ahmed <karahmed@amazon.de>=
 wrote:
>> Add direct access to MSR_IA32_SPEC_CTRL for guests. This is needed for
>> guests
>> that will only mitigate Spectre V2 through IBRS+IBPB and will not be
>> using a
>> retpoline+IBPB based approach.
>>=20
>> To avoid the overhead of atomically saving and restoring the
>> MSR_IA32_SPEC_CTRL
>> for guests that do not actually use the MSR, only add_atomic_switch_msr
>> when a
>> non-zero is written to it.
>=20
>=20
> We tried this and found that it was about 3% slower that doing the old way=
 of rdmsr and wrmsr.
>=20

Do you mean that the host would intercept the guest WRMSR and do WRMSR itsel=
f?  I would suggest that doing so is inconsistent with the docs.  As specifi=
ed, doing WRMSR to write 1 to IBRS does *not* protect the guest.

For that matter, what are the semantics of VMRESUME doing a write to IBRS as=
 part of its MSR switch?  Is it treated as IBRS=3D1 from guest context?