Re: [PATCH v2] KVM: arm64: Don't acquire RCU read lock for exclusive table walks

From: Oliver Upton
Date: Wed Nov 16 2022 - 03:15:31 EST


On Wed, Nov 16, 2022 at 07:27:27AM +0000, Oliver Upton wrote:
> On Wed, Nov 16, 2022 at 03:08:49AM +0000, Marc Zyngier wrote:
> > I'm not crazy about this sort of parameters. I think it would make a
> > lot more sense to pass a pointer to the walker structure and do the
> > flag check inside the helper.
> >
> > That way, we avoid extra churn if/when we need extra state or
> > bookkeeping around the walk.
>
> Sure, let's go that way instead. v3 on the way lol :)

Well, going this route is going to require hoisting around a few things.

I'd very much prefer to keep the RCU indirection ifdeffery all in one
place, but I'll need to haul it all after the definitions of kvm_pgtable_walker
and kvm_pgtable_walk_flags but before the definition of kvm_pgtable
(as it needs kvm_pteref_t).

I'm not too bothered by it, but not quite as small of a bandaid this
time around.

--
Thanks,
Oliver