Re: [PATCH v4 2/4] KVM: x86/mmu: Dynamically allocate shadow MMU's hashed page list

[Sean Christopherson <seanjc@google.com>]

On Wed, May 28, 2025, Sean Christopherson wrote:
> On Wed, May 28, 2025, Xiaoyao Li wrote:
> > On 5/23/2025 8:11 AM, Sean Christopherson wrote:
> > > diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> > > index cbc84c6abc2e..41da2cb1e3f1 100644
> > > --- a/arch/x86/kvm/mmu/mmu.c
> > > +++ b/arch/x86/kvm/mmu/mmu.c
> > > @@ -3882,6 +3882,18 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu *vcpu)
> > >   	return r;
> > >   }
> > > +static int kvm_mmu_alloc_page_hash(struct kvm *kvm)
> > > +{
> > > +	typeof(kvm->arch.mmu_page_hash) h;
> > 
> > Out of curiousity, it is uncommon in KVM to use typeof() given that we know
> > what the type actually is. Is there some specific reason?
> 
> I'm pretty sure it's a leftover from various experiments.  IIRC, I was trying to
> do something odd and was having a hard time getting the type right :-)
> 
> I'll drop the typeof() in favor of "struct hlist_head *", using typeof here isn't
> justified and IMO makes the code a bit harder to read.

Gah, I forgot to switch to address this when applying.  I'll fixup the commit
and force push; it'll only affect this series (hooray for topic branches).