On Wed, Sep 27, 2017 at 02:36:03PM +0200, Paolo Bonzini wrote: > On 26/09/2017 19:12, Josh Triplett wrote: > > Does this make any other checks redundant and removable? > > It would make sense to place it in cpu_has_kvm_support instead cpu_has_kvm_support() or cpu_has_vmx()? >, and the same in svm.c's has_svm. I don't follow (but I also don't know what any of these three letter acryonyms acronyms stand for), does svm depend on vmx or vice-versa?