Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
User-Agent: Cyrus-JMAP/3.5.0-alpha0-1345-g8441cd7852-fm-20211006.001-g8441cd78
Mime-Version: 1.0
Message-Id: <10a8868e-eb85-4043-a5ad-96c03d9b0abb@www.fastmail.com>
In-Reply-To: <df3af1c2-fe93-ea21-56e5-4d70d08e55f2@redhat.com>
References: <20211011215813.558681373@linutronix.de>
 <20211011223611.069324121@linutronix.de>
 <8a5762ab-18d5-56f8-78a6-c722a2f387c5@redhat.com>
 <BYAPR11MB3256B39E2A34A09FF64ECC5BA9B79@BYAPR11MB3256.namprd11.prod.outlook.com>
 <0962c143-2ff9-f157-d258-d16659818e80@redhat.com>
 <BYAPR11MB325676AAA8A0785AF992A2B9A9B79@BYAPR11MB3256.namprd11.prod.outlook.com>
 <da47ba42-b61e-d236-2c1c-9c5504e48091@redhat.com>
 <d673e736-0a72-4549-816d-b755227ea797@www.fastmail.com>
 <df3af1c2-fe93-ea21-56e5-4d70d08e55f2@redhat.com>
Date:   Wed, 13 Oct 2021 07:59:26 -0700
From:   "Andy Lutomirski" <luto@kernel.org>
To:     "Paolo Bonzini" <pbonzini@redhat.com>,
        "Liu, Jing2" <jing2.liu@intel.com>,
        "Thomas Gleixner" <tglx@linutronix.de>,
        "Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>
Cc:     "the arch/x86 maintainers" <x86@kernel.org>,
        "Bae, Chang Seok" <chang.seok.bae@intel.com>,
        "Dave Hansen" <dave.hansen@linux.intel.com>,
        "Arjan van de Ven" <arjan@linux.intel.com>,
        "kvm list" <kvm@vger.kernel.org>,
        "Nakajima, Jun" <jun.nakajima@intel.com>,
        "Jing Liu" <jing2.liu@linux.intel.com>,
        "Sean Christopherson" <seanjc@google.com>
Subject: Re: [patch 13/31] x86/fpu: Move KVMs FPU swapping to FPU core
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk


On Wed, Oct 13, 2021, at 5:26 AM, Paolo Bonzini wrote:
> On 13/10/21 12:14, Andy Lutomirski wrote:
>>> I think it's simpler to always wait for #NM, it will only happen
>>> once per vCPU.  In other words, even if the guest clears XFD before
>>> it generates #NM, the guest_fpu's XFD remains nonzero and an #NM
>>> vmexit is possible.  After #NM the guest_fpu's XFD is zero; then
>>> passthrough can happen and the #NM vmexit trap can be disabled.
>>
>> This will stop being at all optimal when Intel inevitably adds
>> another feature that uses XFD.  In the potentially infinite window in
>> which the guest manages XFD and #NM on behalf of its userspace and
>> when the guest allocates the other hypothetical feature, all the #NMs
>> will have to be trapped by KVM.
>
> The reason is that it's quite common to simply let the guest see all=20
> CPUID bits that KVM knows about.  But it's not unlikely that most gues=
ts=20
> will not ever use any XFD feature, and therefore will not ever see an=20
> #NM.  I wouldn't have any problem with allocating _all_ of the dynamic=20
> state space on the first #NM.
>
> Thinking more about it, #NM only has to be trapped if XCR0 enables a=20
> dynamic feature.  In other words, the guest value of XFD can be limite=
d=20
> to (host_XFD|guest_XFD) & guest_XCR0.  This avoids that KVM=20
> unnecessarily traps for old guests that use CR0.TS.
>

You could simplify this by allocating the state the first time XCR0 enab=
les the feature in question.

(This is how regular non-virt userspace *should* work too, but it looks =
like I=E2=80=99ve probably been outvoted on that front=E2=80=A6)

> Paolo
>
>> Is it really worthwhile for KVM to use XFD at all instead of
>> preallocating the state and being done with it?  KVM would still have
>> to avoid data loss if the guest sets XFD with non-init state, but #NM
>> could always pass through.
>>