Re: [PATCH 1/5] Implement #NMI exiting for nested SVM

[Jan Kiszka <jan.kiszka@siemens.com>]

Alexander Graf wrote:
> When injecting an NMI to the l1 guest while it was running the l2 guest, we
> didn't #VMEXIT but just injected the NMI to the l2 guest.
> 
> Let's be closer to real hardware and #VMEXIT if we're supposed to do so.
> 
> Signed-off-by: Alexander Graf <agraf@suse.de>
> ---
>  arch/x86/kvm/svm.c |   38 ++++++++++++++++++++++++++++++++------
>  1 files changed, 32 insertions(+), 6 deletions(-)
> 
> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
> index 9a4daca..f12a669 100644
> --- a/arch/x86/kvm/svm.c
> +++ b/arch/x86/kvm/svm.c
> @@ -1375,6 +1375,21 @@ static int nested_svm_check_exception(struct vcpu_svm *svm, unsigned nr,
>  	return nested_svm_exit_handled(svm);
>  }
>  
> +static inline int nested_svm_nmi(struct vcpu_svm *svm)
> +{
> +	if (!is_nested(svm))
> +		return 0;
> +
> +	svm->vmcb->control.exit_code = SVM_EXIT_NMI;
> +
> +	if (nested_svm_exit_handled(svm)) {
> +		nsvm_printk("VMexit -> NMI\n");
> +		return 1;
> +	}
> +
> +	return 0;
> +}
> +
>  static inline int nested_svm_intr(struct vcpu_svm *svm)
>  {
>  	if (!is_nested(svm))
> @@ -2462,7 +2477,9 @@ static int svm_nmi_allowed(struct kvm_vcpu *vcpu)
>  	struct vcpu_svm *svm = to_svm(vcpu);
>  	struct vmcb *vmcb = svm->vmcb;
>  	return !(vmcb->control.int_state & SVM_INTERRUPT_SHADOW_MASK) &&
> -		!(svm->vcpu.arch.hflags & HF_NMI_MASK);
> +		!(svm->vcpu.arch.hflags & HF_NMI_MASK) &&
> +		gif_set(svm) &&
                ^^^^^^^^^^^^
I'm not claiming to be up-to-date with the SVM code around NMI
injection, but this addition irritates me. Can you explain why I don't
have to worry that a cleared IF could now defer NMI injections for L1
guests?

> +		!is_nested(svm);
>  }
>  
>  static int svm_interrupt_allowed(struct kvm_vcpu *vcpu)
> @@ -2488,22 +2505,31 @@ static void enable_irq_window(struct kvm_vcpu *vcpu)
>  	struct vcpu_svm *svm = to_svm(vcpu);
>  	nsvm_printk("Trying to open IRQ window\n");
>  
> -	nested_svm_intr(svm);
> +	if (nested_svm_intr(svm))
> +		return;
>  
>  	/* In case GIF=0 we can't rely on the CPU to tell us when
>  	 * GIF becomes 1, because that's a separate STGI/VMRUN intercept.
>  	 * The next time we get that intercept, this function will be
>  	 * called again though and we'll get the vintr intercept. */
> -	if (gif_set(svm)) {
> -		svm_set_vintr(svm);
> -		svm_inject_irq(svm, 0x0);
> -	}
> +	if (!gif_set(svm))
> +		return;
> +
> +	svm_set_vintr(svm);
> +	svm_inject_irq(svm, 0x0);

The last change is pure refactoring that should not belong into this
patch, should it?

>  }
>  
>  static void enable_nmi_window(struct kvm_vcpu *vcpu)
>  {
>  	struct vcpu_svm *svm = to_svm(vcpu);
>  
> +	if (nested_svm_nmi(svm))
> +		return;
> +
> +	/* NMI is deferred until GIF == 1. Setting GIF will cause a #VMEXIT */
> +	if (!gif_set(svm))
> +		return;
> +

The second half is an unrelated optimization? Then please file a
separate patch.

>  	if ((svm->vcpu.arch.hflags & (HF_NMI_MASK | HF_IRET_MASK))
>  	    == HF_NMI_MASK)
>  		return; /* IRET will cause a vm exit */

Jan

-- 
Siemens AG, Corporate Technology, CT SE 2
Corporate Competence Center Embedded Linux