[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <c7d87723-5533-257f-fbf0-ecd3a0b96602@redhat.com>
Date: Wed, 17 Nov 2021 13:53:42 +0100
From: Paolo Bonzini <pbonzini@...hat.com>
To: "Tian, Kevin" <kevin.tian@...el.com>,
"Nakajima, Jun" <jun.nakajima@...el.com>,
Thomas Gleixner <tglx@...utronix.de>
Cc: "Liu, Jing2" <jing2.liu@...el.com>,
LKML <linux-kernel@...r.kernel.org>,
"x86@...nel.org" <x86@...nel.org>,
"kvm@...r.kernel.org" <kvm@...r.kernel.org>,
Dave Hansen <dave.hansen@...ux.intel.com>,
Arjan van de Ven <arjan@...ux.intel.com>,
Jing Liu <jing2.liu@...ux.intel.com>,
"seanjc@...gle.com" <seanjc@...gle.com>,
"Cooper, Andrew" <andrew.cooper3@...rix.com>,
"Bae, Chang Seok" <chang.seok.bae@...el.com>
Subject: Re: Thoughts of AMX KVM support based on latest kernel
On 11/17/21 11:15, Tian, Kevin wrote:
> We are not sure whether such trick is worthwhile, since a sane
> guest shouldn't set XFD[AMX]=1 before storing the AMX state. This
> is why we want to seek SDM change to mark out that the software
> should not assume XTILEDATA is still valid when XFD[AMX]=1.
Okay, I just don't want it to be called out as virtualization specific.
It doesn't have to happen in current processors, but it should be
architecturally valid behavior to clear the processor's state as soon as
a bit in XFD is set to 1.
Paolo
Powered by blists - more mailing lists