On 08/02/2018 16:54, Jim Mattson wrote: > On Thu, Feb 8, 2018 at 7:29 AM, Jim Mattson <jmattson@xxxxxxxxxx> wrote: >> Similarly, the correct L1 CR4 value should be in vmcs01's CR4 >> read shadow field. > Sorry; that's wrong. L1's CR4 value has to be reconstructed from the > vmcs01 guest CR4 field and CR4 shadow field using the cr4 guest/host > mask. But there is no way to get it from any field(s) in vmcs12. Now that we have the prepare_vmcs02_full/prepare_vmcs02 split, we probably should do more checks in there, and not rely on the processor anymore. Paolo
