Re: [PATCH v3 2/5] KVM: MMU: check guest CR3 reserved bits based on its physical address width.

[Yu Zhang <yu.c.zhang@linux.intel.com>]

On 9/18/2017 4:41 PM, Paolo Bonzini wrote:
> On 18/09/2017 10:15, Yu Zhang wrote:
>>> static bool emulator_get_cpuid(struct x86_emulate_ctxt *ctxt,
>>>                           u32 *eax, u32 *ebx, u32 *ecx, u32 *edx, bool
>>> check_limit)
>>> {
>>>           return kvm_cpuid(emul_to_vcpu(ctxt), eax, ebx, ecx, edx,
>>> check_limit);
>>> }
>>>
>>> And:
>>>
>>> bool kvm_cpuid(struct kvm_vcpu *vcpu, u32 *eax, u32 *ebx,
>>>          u32 *ecx, u32 *edx, bool check_limit)
>>> {
>>> u32 function = *eax, index = *ecx;
>>> struct kvm_cpuid_entry2 *best;
>>> bool entry_found = true;
>>> ...
>>>
>>> Doesn't this immediately try to dereference a NULL pointer?  How much
>>> testing have you done of this code?
>> Thanks Jim.
>> I have tested this code in a simulator to successfully boot a VM in
>> shadow mode. Seems this code is not covered(but I am now still
>> perplexed why this is not covered). Any possibility that the
>> check_cr_write() is not triggered when emulating the cr operations?
> CR moves usually don't go through the emulator (the main exception is
> emulation of invalid guest state when the processor doesn't support
> unrestricted_guest=1, but even that is unlikely to happen with
> EFER.LMA=1).  This explains why you didn't see the failure.

Oh, right. It normally goes to handle_cr(). Thanks, Paolo.

Yu

>
>> Anyway, this should be a bug and thanks for pointing this out, and
>> I'll send out the fix later.
> Thanks,
>
> Paolo
>