On Thu, Sep 30, 2021, Paolo Bonzini wrote: > On 30/09/21 00:24, Sean Christopherson wrote: > > * RESET since KVM emulates RESET before exposing the vCPU to userspace, > > * i.e. it'simpossible for kvm_cpuid() to find a valid entry on RESET. > > + * But, go through the motions in case that's ever remedied. Note, the > > + * index for CPUID.0x1 is not significant, arbitrarily specify '0'. > > Just one nit, this comment change is not really needed because almost all > callers are using '0' for the same reason. > > But, perhaps adding kvm_find_cpuid_entry_index and removing the last > parameter from kvm_find_cpuid_entry would be a good idea. I like this idea, but only if callers are forced to specify the index when the index is significant, e.g. add a magic CPUID_INDEX_DONT_CARE and WARN in cpuid_entry2_find() if index is significant and index == DONT_CARE. I'll fiddle with this, unless you want the honors? > Also, the kvm_cpuid() reference needs to be changed, which I did upon > commit. Doh, thanks!
