Re: [RESEND PATCH 5/6] KVM: x86/VMX: add kvm_vmx_reinject_nmi_irq() for NMI/IRQ reinjection

[Peter Zijlstra]

On Fri, Nov 11, 2022 at 01:04:27PM +0100, Paolo Bonzini wrote:
> On 11/11/22 10:15, Peter Zijlstra wrote:
> > I don't speak virt (but this all sounds disguisting)
> 
> Yes, it is.  AMD does not need this, they just hold onto the interrupt
> until the host has issued both STGI (for NMIs) and STI (for IRQs).

Right -- Cooper just explained that to me.

> On Intel you can optionally make it hold onto IRQs, but NMIs are always
> eaten by the VMEXIT and have to be reinjected manually.

That 'optionally' thing worries me -- as in, KVM is currently
opting-out?

Since much of this is about preparing for FRED, can we pretty *PLEASE*
fix this VMX NMI hole instead so that all this nonsense isn't needed
anymore, please, really please?

Have FRED imply a GI flag for VMX or something and then we can all
forget about these reinjection horrors.