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-alpha0-1118-g75eff666e5-fm-20210816.002-g75eff666
Mime-Version: 1.0
Message-Id: <ea7eddc4-decd-47b7-b98b-c04c441495d9@www.fastmail.com>
In-Reply-To: <3399412.qF98CnctbS@tjmaciei-mobl5>
References: <20210730145957.7927-1-chang.seok.bae@intel.com>
 <YR1HYRRN0HMTxXrw@zn.tnic> <BCC327C2-CF9F-4910-B626-315E515E9A3A@intel.com>
 <3399412.qF98CnctbS@tjmaciei-mobl5>
Date:   Wed, 18 Aug 2021 18:21:11 -0700
From:   "Andy Lutomirski" <luto@kernel.org>
To:     "Thiago Macieira" <thiago.macieira@intel.com>,
        "Borislav Petkov" <bp@alien8.de>,
        "Bae, Chang Seok" <chang.seok.bae@intel.com>
Cc:     "Thomas Gleixner" <tglx@linutronix.de>,
        "Ingo Molnar" <mingo@kernel.org>,
        "the arch/x86 maintainers" <x86@kernel.org>,
        "Brown, Len" <len.brown@intel.com>,
        "Dave Hansen" <dave.hansen@intel.com>,
        "Liu, Jing2" <jing2.liu@intel.com>,
        "Shankar, Ravi V" <ravi.v.shankar@intel.com>,
        "Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>
Subject: =?UTF-8?Q?Re:_[PATCH_v9_12/26]_x86/fpu/xstate:_Use_feature_disable_(XFD)?=
 =?UTF-8?Q?_to_protect_dynamic_user_state?=
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk


On Wed, Aug 18, 2021, at 2:04 PM, Thiago Macieira wrote:
> On Wednesday, 18 August 2021 13:43:50 PDT Bae, Chang Seok wrote:
> > > Then our API needs improving. An app should be able to ask the ker=
nel
> > > "Do you support AMX?" get a proper answer and act accordingly.
> >=20
> > Maybe I=E2=80=99m missing something, but I wonder what=E2=80=99s the=
 difference from
> > reading  XCR0.
>=20
> That assumes the kernel will always enable the bits in XCR0, like it i=
s doing=20
> today and with your patch, because modifying it is a VM exit.
>=20
> But it's not the only possible solution. A future kernel could decide =
to leave=20
> some bits off and only enable upon request. That's how macOS/Darwin do=
es its=20
> AVX512 support.

The fact that Darwin does this strongly suggests that real programs can =
handle it, which increases my inclination for Linux to do the same thing.

>=20
> --=20
> Thiago Macieira - thiago.macieira (AT) intel.com
>   Software Architect - Intel DPG Cloud Engineering
>=20
>=20
>=20
>=20