Re: [PATCH] KVM: x86: Use a stable condition around all VT-d PI paths

[Paolo Bonzini]

On 11/16/21 18:42, Sean Christopherson wrote:
> > +	return kvm_arch_has_assigned_device(kvm) &&
> > +		irq_remapping_cap(IRQ_POSTING_CAP) &&
> > +		irqchip_in_kernel(kvm) && enable_apicv;
> Bad indentation/alignment.

What is even the right indentation?  I'd just wrap everything in 
parentheses but then check patch complains "return is not a function" 
(NSS), so I went for two tabs and called it a day.

> Not that it's likely to matter, but would it make sense to invert the checks so
> that they're short-circuited on the faster KVM checks?  E.g. fastest to slowest:
>
> 	return irqchip_in_kernel(kvm) && enable_apic &&
> 	       kvm_arch_has_assigned_device(kvm) &&
> 	       irq_remapping_cap(IRQ_POSTING_CAP);

Sure, why not.

Paolo