Re: [PATCH v2 4/4] KVM: x86: Add VCPU substate for NMI states

[Marcelo Tosatti <mtosatti@redhat.com>]

On Thu, Oct 15, 2009 at 07:05:36PM +0200, Jan Kiszka wrote:
> This plugs an NMI-related hole in the VCPU synchronization between
> kernel and user space. So far, neither pending NMIs nor the inhibit NMI
> mask was properly read/set which was able to cause problems on
> vmsave/restore, live migration and system reset. Fix it by making use
> of the new VCPU substate interface.
> 
> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
> ---
> 
>  Documentation/kvm/api.txt       |   12 ++++++++++++
>  arch/x86/include/asm/kvm.h      |    7 +++++++
>  arch/x86/include/asm/kvm_host.h |    2 ++
>  arch/x86/kvm/svm.c              |   22 ++++++++++++++++++++++
>  arch/x86/kvm/vmx.c              |   30 ++++++++++++++++++++++++++++++
>  arch/x86/kvm/x86.c              |   26 ++++++++++++++++++++++++++
>  6 files changed, 99 insertions(+), 0 deletions(-)
> 
> diff --git a/Documentation/kvm/api.txt b/Documentation/kvm/api.txt
> index bee5bbd..e483edb 100644
> --- a/Documentation/kvm/api.txt
> +++ b/Documentation/kvm/api.txt
> @@ -848,3 +848,15 @@ Deprecates: KVM_GET/SET_CPUID2
>  Architectures: x86
>  Payload: struct kvm_lapic
>  Deprecates: KVM_GET/SET_LAPIC
> +
> +6.8 KVM_X86_VCPU_STATE_NMI
> +
> +Architectures: x86
> +Payload: struct kvm_nmi_state
> +Deprecates: -
> +
> +struct kvm_nmi_state {
> +       __u8 pending;
> +       __u8 masked;
> +       __u8 pad1[6];

Don't you also have to save "nmi_injected", in case of failure during
NMI delivery.

BTW, what happens to exceptions that fail to be delivered? Can't see
where they are saved/restored across migration.