Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 16 Nov 2022 20:22:23 +0000
From:   Sean Christopherson <seanjc@google.com>
To:     Borislav Petkov <bp@alien8.de>
Cc:     Kim Phillips <kim.phillips@amd.com>, x86@kernel.org,
        Boris Ostrovsky <boris.ostrovsky@oracle.com>,
        Dave Hansen <dave.hansen@linux.intel.com>,
        "H. Peter Anvin" <hpa@zytor.com>, Ingo Molnar <mingo@redhat.com>,
        Joao Martins <joao.m.martins@oracle.com>,
        Jonathan Corbet <corbet@lwn.net>,
        Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>,
        Paolo Bonzini <pbonzini@redhat.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        David Woodhouse <dwmw@amazon.co.uk>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Juergen Gross <jgross@suse.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Tony Luck <tony.luck@intel.com>,
        Babu Moger <Babu.Moger@amd.com>,
        Tom Lendacky <thomas.lendacky@amd.com>, kvm@vger.kernel.org,
        linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/3] x86/cpufeatures: Add support for cpuid leaf
 80000021/EAX (FeatureExt2Eax)
Message-ID: <Y3VGf8WsvxZ/S1aI@google.com>
References: <20221104213651.141057-1-kim.phillips@amd.com>
 <20221104213651.141057-2-kim.phillips@amd.com>
 <Y2WIy2A1RuQE/9SK@zn.tnic>
 <c00b1a65-c885-c874-79cb-16011ac82eb3@amd.com>
 <Y3TQsUmTieC4NnO/@zn.tnic>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Y3TQsUmTieC4NnO/@zn.tnic>
Precedence: bulk

On Wed, Nov 16, 2022, Borislav Petkov wrote:
> On Tue, Nov 15, 2022 at 05:10:50PM -0600, Kim Phillips wrote:
> > When trying to wire up a scattered host AUTOIBRS version up to
> > kvm, I couldn't get past all the reverse_cpuid_check()
> > BUILD_BUGs demanding exclusivity between h/w and "Linux"
> > (s/w) FEATUREs.

FWIW, it's not exclusivity per se, it's to ensure that any CPUID bits KVM wants
to advertise to userspace uses the architectural definition and not the kernel's
software defined info.  This allows KVM to do things like

	if (guest_cpuid_has(X86_FEATURE_AUTOIBRS))

and guarantee that the lookup on guest CPUID, which follows the architectural
layout, will look at the correct leaf+subleaf+reg+bit.

> I guess something like below.
> 
> Sean, can you pls check the KVM bits whether I've done them all right?

Looks correct.