On Mon, Mar 25, 2024 at 09:43:31PM +1300, "Huang, Kai" <kai.huang@xxxxxxxxx> wrote: > > > > Currently, the KVM x86 always reports KVM_MAX_VCPUS for all VMs but doesn't > > > allow to enable KVM_CAP_MAX_VCPUS to configure the number of maximum vcpus > > maximum number of vcpus > > > on VM-basis. > > > > > > Add "per-VM maximum vcpus" to KVM x86/TDX to accommodate TDX's needs. > > > > > > The userspace-configured value then can be verified when KVM is actually > > used > > > creating the TDX guest. > > > " > > I think we still have two options regarding to how 'max_vcpus' is handled in > ioctl() to do TDH.MNG.INIT: > > 1) Just use the 'max_vcpus' done in KVM_ENABLE_CAP(KVM_CAP_MAX_VCPUS), > 2) Still pass the 'max_vcpus' as input, but KVM verifies it against the > value that is saved in KVM_ENABLE_CAP(KVM_CAP_MAX_VCPUS). > > 2) seems unnecessary, so I don't have objection to use 1). But it seems we > could still mention it in the changelog in that patch? Sure, let me update the commit log. -- Isaku Yamahata <isaku.yamahata@xxxxxxxxx>
