[PATCH v3 3/3] arm64: Add workaround for Arm Cortex-A77 erratum 1508412

[Will Deacon]

On Fri, Jul 31, 2020 at 04:55:26PM -0600, Rob Herring wrote:
> On Thu, Jul 30, 2020 at 2:22 AM Will Deacon <will at kernel.org> wrote:
> >
> > On Wed, Jul 29, 2020 at 05:38:00PM +0100, Catalin Marinas wrote:
> > > On Fri, Jul 17, 2020 at 02:52:33PM -0600, Rob Herring wrote:
> > > > diff --git a/arch/arm64/include/asm/kvm_hyp.h b/arch/arm64/include/asm/kvm_hyp.h
> > > > index ce3080834bfa..ce5b0d9b12bf 100644
> > > > --- a/arch/arm64/include/asm/kvm_hyp.h
> > > > +++ b/arch/arm64/include/asm/kvm_hyp.h
> > > > @@ -46,6 +46,17 @@
> > > >  #define read_sysreg_el2(r) read_sysreg_elx(r, _EL2, _EL1)
> > > >  #define write_sysreg_el2(v,r)      write_sysreg_elx(v, r, _EL2, _EL1)
> > > >
> > > > +static inline u64 __hyp_text read_sysreg_par(void)
> > > > +{
> > > > +   u64 par;
> > > > +   if (cpus_have_const_cap(ARM64_WORKAROUND_1508412))
> > > > +           dmb(sy);
> > > > +   par = read_sysreg(par_el1);
> > > > +   if (cpus_have_const_cap(ARM64_WORKAROUND_1508412))
> > > > +           dmb(sy);
> > > > +   return par;
> > > > +}
> > >
> > > Even if that's not always called on a critical path, I agree with Andrew
> > > that we could use alternatives here for dmb(sy).
> >
> > Even then, I'm not sure how this helps if the CPU can speculatively branch
> > to the PAR access without executing the DMB. Is that not possible on A77?
> 
> I'm told by the h/w folks speculation is not possible in this case.

Thanks. Could you add a comment to that effect, please?

Will