Date:   Tue, 28 Feb 2023 16:40:38 -0800
From:   Isaku Yamahata <isaku.yamahata@gmail.com>
To:     "Huang, Kai" <kai.huang@intel.com>
Cc:     "isaku.yamahata@gmail.com" <isaku.yamahata@gmail.com>,
        "Christopherson,, Sean" <seanjc@google.com>,
        "Shahar, Sagi" <sagis@google.com>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "Aktas, Erdem" <erdemaktas@google.com>,
        "pbonzini@redhat.com" <pbonzini@redhat.com>,
        "kvm@vger.kernel.org" <kvm@vger.kernel.org>,
        "Yamahata, Isaku" <isaku.yamahata@intel.com>,
        "dmatlack@google.com" <dmatlack@google.com>
Subject: Re: [PATCH v11 017/113] KVM: Support KVM_CAP_MAX_VCPUS for
 KVM_ENABLE_CAP
Message-ID: <20230301004038.GB3852341@ls.amr.corp.intel.com>
References: <cover.1673539699.git.isaku.yamahata@intel.com>
 <01e061779b88ce4d32bbe483ed2bd3224cd8e330.1673539699.git.isaku.yamahata@intel.com>
 <5f9870acec07c47de5bab460c8c0d4a2c0db08fc.camel@intel.com>
 <20230227212605.GF4175971@ls.amr.corp.intel.com>
 <bdf8b2e1aa52ae8f4d5e2a6c92e0f8e426b97234.camel@intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <bdf8b2e1aa52ae8f4d5e2a6c92e0f8e426b97234.camel@intel.com>
Precedence: bulk

On Tue, Feb 28, 2023 at 09:57:50PM +0000,
"Huang, Kai" <kai.huang@intel.com> wrote:

> On Mon, 2023-02-27 at 13:26 -0800, Isaku Yamahata wrote:
> > > > TDX attestation includes the maximum number of vcpu that the guest can
> > > > accommodate.  
> > > > 
> > > 
> > > I don't understand why "attestation" is the reason here.  Let's say TDX is
> > > used
> > > w/o attestation, I don't think this patch can be discarded?
> > > 
> > > IMHO the true reason is TDX has it's own control of maximum number of vcpus,
> > > i.e. asking you to specify the value when creating the TD.  Therefore, the
> > > constant KVM_MAX_VCPUS doesn't work for TDX guest anymore.
> > 
> > Without TDX attestation, this can be discarded.  The TD is created with
> > max_vcpus=KVM_MAX_VCPUS by default.
> 
> This parses like: 
> 
> If we have attestation, the TD can be created with a user-specified non-default
> value.  Otherwise, the TD is always created with default value.
> 
> It doesn't make sense, right?
> 
> Because architecturally whether TD can be created with a user specified value
> doesn't depend on attestation at all.

I'm not sure if I got your point.
Even without attestation, it's allowed to specify max vcpus.   Not "always".
-- 
Isaku Yamahata <isaku.yamahata@gmail.com>