Re: [PATCH] x86/kvm: Handle async page faults directly through do_page_fault()

[Paolo Bonzini <pbonzini@redhat.com>]

On 28/02/20 20:04, Andy Lutomirski wrote:
>>> +      * We are relying on the interrupted context being sane (valid
>>> +      * RSP, relevant locks not held, etc.), which is fine as long as
>>> +      * the the interrupted context had IF=1.
>> This is not about IF=0/IF=1; the KVM code is careful about taking
>> spinlocks only with IRQs disabled, and async PF is not delivered if the
>> interrupted context had IF=0.  The problem is that the memory location
>> is not reentrant if an NMI is delivered in the wrong window, as you hint
>> below.
>
> If an async PF is delivered with IF=0, then, unless something else
> clever happens to make it safe, we are toast. 

Right, it just cannot happen.  kvm_can_do_async_pf is where KVM decides
whether a page fault must be handled synchronously, and it does this:

bool kvm_can_do_async_pf(struct kvm_vcpu *vcpu)
{
	...
        /*
         * If interrupts are off we cannot even use an artificial
         * halt state.
         */
        return kvm_x86_ops->interrupt_allowed(vcpu);
}

The same function is called by kvm_arch_can_inject_async_page_present.

Paolo