On Thu, Aug 23, 2012 at 05:24:00PM +0900, Takuya Yoshikawa wrote:
> On Thu, 23 Aug 2012 15:42:49 +0800
> Gavin Shan <shangw@xxxxxxxxxxxxxxxxxx> wrote:
>
> > The build error was caused by that builtin functions are calling
> > the functions implemented in modules. That was introduced by the
> > following commit.
> >
> > commit 4d8b81abc47b83a1939e59df2fdb0e98dfe0eedd
> >
> > The patches fix that to convert the gfn to hva in direct way.
>
> Not just that...
>
> >
> > Signed-off-by: Gavin Shan <shangw@xxxxxxxxxxxxxxxxxx>
> > ---
> > diff --git a/arch/powerpc/kvm/book3s_hv_rm_mmu.c b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
> > index 56ac1a5..0958523 100644
> > --- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c
> > +++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
> > @@ -197,7 +197,8 @@ long kvmppc_h_enter(struct kvm_vcpu *vcpu, unsigned long flags,
> > pa &= PAGE_MASK;
> > } else {
> > /* Translate to host virtual address */
> > - hva = gfn_to_hva_memslot(memslot, gfn);
> > + hva = memslot->userspace_addr +
> > + (gfn - memslot->base_gfn) * PAGE_SIZE;
>
> Although I do not think what that commit wanted to do was wrong,
> I believe the gfn_to_hva_memslot() change should have been
> explained more explicitely, at least, to all users of
> the not-anymore-inline function.
I definitely agree. The patch description on 4d8b81abc4 is woefully
inadequate.
> Probably kvm-ppc people should re-check the call sites: what they
> want to use may be newly introduced __gfn_to_hva_memslot() which
> is equivalent to the original gfn_to_hva_memslot().
Yes, I think we should move __gfn_to_hva_memslot to kvm_host.h and
make it static inline, and use it in kvmppc_h_enter().
Paul.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
