Re: [PATCH v2 1/7] KVM:VMX: Define CET VMCS fields and bits

[Paolo Bonzini]

On 29/01/19 09:29, Yang Weijiang wrote:
> On Tue, Jan 29, 2019 at 04:19:34PM +0100, Paolo Bonzini wrote:
>> On 28/01/19 11:33, Yang Weijiang wrote:
>>>> There is no code in this series to pass these fields to and from
>>>> userspace, and also to save/restore U_CET, INT_SSP_TAB, PL0_SSP and
>>>> PL3_SSP across context switches.
>>>>
>>> The kernel consumes these MSRs, please see kernel CET patch:
>>> https://lkml.org/lkml/fancy/2018/11/20/225
>>
>> Still, even if the kernel saves these fields across context switch in
>> XSAVE areas, KVM must support accesses to the MSRs from userspace, for
>> example in order to perform live migration.
>>
>> For example, when reading/writing these in kvm_set_msr or
>> kvm_get_msr_common, you would have to do a read/write from the host
>> MSRs.  You also have to put kvm_load_guest_fpu/kvm_put_guest_fpu calls
>> in __msr_io.
> Thanks, you're right, if we want to support live migration we need do that.
> However, this feature relies on quite a few external
> dependencies, e.g., cpu capabilities, kernel version and configuration, we
> don't support live migration right now. what's your opinion?

Live migration is generally a requirement for a feature to be accepted
in KVM.

Paolo

>> Thanks,
>>
>> Paolo
>>
>>>> In addition, PL1_SSP and PL2_SSP should be supported even if the guest
>>>> doesn't use them.  It makes sense to avoid intercepting them, but they
>>>> should still be supported and switched (possibly only if nonzero).
>>>>
>>>> Am I missing something, for example a dependency on host CET support?
>>>> If not, how was this series tested?
>>>>
>>> The guest CET feature is tested with kernel CET patches on internal
>>> virtual platform.
>>>