On Thu, 17 Nov 2022 at 03:14, Richard Henderson
<richard.hender...@linaro.org> wrote:
>
> On 11/16/22 09:03, Ard Biesheuvel wrote:
> > diff --git a/target/arm/ptw.c b/target/arm/ptw.c
> > index 3745ac9723474332..9a6277d862fac229 100644
> > --- a/target/arm/ptw.c
> > +++ b/target/arm/ptw.c
> > @@ -1222,6 +1222,14 @@ static bool get_phys_addr_lpae(CPUARMState *env,
> > S1Translate *ptw,
> > ps = MIN(ps, param.ps);
> > assert(ps < ARRAY_SIZE(pamax_map));
> > outputsize = pamax_map[ps];
> > +
> > + /*
> > + * With LPA2, the effective output address (OA) size is at most 48
> > bits
> > + * unless TCR.DS == 1
> > + */
> > + if (!param.ds && param.gran != Gran64K) {
> > + outputsize = MIN(outputsize, 48);
> > + }
>
> Reviewed-by: Richard Henderson <richard.hender...@linaro.org>
>
> I thought about moving this back into aa64_va_parameters, similar to how we
> bound tsz, but
> since this is the only use of param.ps, this placement is as good as any.
Applied to target-arm.next, thanks.
-- PMM
