[RFC PATCH 16/27] KVM: arm64: Prepare Hyp memory protection

[Quentin Perret]

On Monday 07 Dec 2020 at 11:16:05 (+0000), Fuad Tabba wrote:
> On Fri, Dec 4, 2020 at 6:01 PM Quentin Perret <qperret at google.com> wrote:
> >
> > On Thursday 03 Dec 2020 at 12:57:33 (+0000), Fuad Tabba wrote:
> > <snip>
> > > > +int hyp_create_idmap(void);
> > > > +int hyp_map_vectors(void);
> > > > +int hyp_back_vmemmap(phys_addr_t phys, unsigned long size, phys_addr_t back);
> > > > +int hyp_cpu_set_vector(enum arm64_hyp_spectre_vector slot);
> > > > +int hyp_create_mappings(void *from, void *to, enum kvm_pgtable_prot prot);
> > > > +int __hyp_create_mappings(unsigned long start, unsigned long size,
> > > > +                         unsigned long phys, unsigned long prot);
> > > > +unsigned long __hyp_create_private_mapping(phys_addr_t phys, size_t size,
> > > > +                                          unsigned long prot);
> > > > +
> > >
> > > nit: I also thought that the hyp_create_mappings function names are a
> > > bit confusing, since there's the create_hyp_mappings functions which
> > > use the aforementioned *hyp_pgtable.
> >
> > Sure, happy to re-name those (and hyp_pgtable above). Any suggestions?
> 
> Perhaps something to indicate that these are temporary, tmp_ or
> bootstrap_ maybe?

Hmm, the thing is these are temporary only in protected mode, they're
permanent otherwise :/

Perhaps I could prefix the protected pgtable (and associated functions)
with 'pkvm_' or so? Marc, any preferences?

Thanks,
Quentin