Re: [PATCH v6 00/29] KVM: Scalable memslots implementation

[Sean Christopherson]

On Tue, Nov 30, 2021, Maciej S. Szmigiero wrote:
> From: "Maciej S. Szmigiero" <maciej.szmigiero@xxxxxxxxxx>
> 
> This series contains the sixth iteration of the scalable memslots patch set.
> It is based on Sean's version "5.5", but with integrated patches for issues
> that arose during its review round.
> 
> In addition to that, the kvm_for_each_memslot_in_gfn_range() implementation
> was reworked to return only strictly overlapping memslots and to use
> iterators.
> 
> However, I've dropped a similar kvm_for_each_memslot_in_hva_range() rework
> since the existing implementation was already returning only strictly
> overlapping memslots and it was already based on interval tree iterators,
> so wrapping them in another layer of iterators would only add unnecessary
> complexity.
> The code in this "for"-like macro is also self-contained and very simple,
> so let's keep it this way.

If kvm_for_each_memslot_in_hva_range() ever gains a user outside of kvm_main.c
it should definitely get an iterator container so that callers don't need to do
the container_of() stuff.  I'd still prefer a container even now, but it's not a
sticking point.