Re: [PATCH v19 037/130] KVM: TDX: Make KVM_CAP_MAX_VCPUS backend specific

[Huang, Kai]

On Thu, 2024-05-30 at 16:12 -0700, Sean Christopherson wrote:
> On Thu, May 30, 2024, Kai Huang wrote:
> > On Wed, 2024-05-29 at 16:15 -0700, Sean Christopherson wrote:
> > > In the unlikely event there is a legitimate reason for max_vcpus_per_td being
> > > less than KVM's minimum, then we can update KVM's minimum as needed.  But AFAICT,
> > > that's purely theoretical at this point, i.e. this is all much ado about nothing.
> > 
> > I am afraid we already have a legitimate case: TD partitioning.  Isaku
> > told me the 'max_vcpus_per_td' is lowed to 512 for the modules with TD
> > partitioning supported.  And again this is static, i.e., doesn't require
> > TD partitioning to be opt-in to low to 512.
> 
> So what's Intel's plan for use cases that creates TDs with >512 vCPUs?

I don't think we have such use cases.  Let me double check with TDX module
guys.  


> 
> > So AFAICT this isn't a theoretical thing now.
> > 
> > Also, I want to say I was wrong about "MAX_VCPUS" in the TD_PARAMS is part
> > of attestation.  It is not.  TDREPORT dosen't include the "MAX_VCPUS", and
> > it is not involved in the calculation of the measurement of the guest.
> > 
> > Given "MAX_VCPUS" is not part of attestation, I think there's no need to
> > allow user to change kvm->max_vcpus by enabling KVM_ENABLE_CAP ioctl() for
> > KVM_CAP_MAX_VCPUS.
> 
> Sure, but KVM would still need to advertise the reduced value for KVM_CAP_MAX_VCPUS
> when queried via KVM_CHECK_EXTENSION.  And userspace needs to be conditioned to
> do a VM-scoped check, not a system-scoped check.

Oh yes.