On Thu, 2025-01-23 at 13:35 +0100, Vitaly Kuznetsov wrote: > > I guess we can change the logic the following: when KVM_SET_CPUID2 is > called on a vCPU again we check that all entries which KVM did not touch > match. For that, we will need to keep a list of mangled entries so we > can introduce a kvm_mangle_cpuid_entry() helper to avoid the need to > keep a static list. Personally, I'm not sure this is not an overkill > though. Putting that another way, KVM would exempt the dynamic entries that KVM is going to overwrite by *itself*, from the comparison. By keeping a list of the entries that it's going to overwrite. I'm not sure I'd call that overkill. I think I prefer it to the option of mangling the CPUID at *runtime* in kvm_cpuid() while the entries in the array differ from what the guest actually sees.
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
