Re: [Xen-devel] [PATCH 7/9] xen/arm: traps: MMIO should only be emulated for fault translation

[Julien Grall <julien.grall@xxxxxxx>]

On 14/07/16 16:28, Stefano Stabellini wrote:
> On Thu, 14 Jul 2016, Julien Grall wrote:
> > Hi Stefano,
> >
> > On 14/07/16 16:06, Stefano Stabellini wrote:
> > > On Wed, 22 Jun 2016, Julien Grall wrote:
> > > > -    if (handle_mmio(&info))
> > > > -    {
> > > > -        advance_pc(regs, hsr);
> > > > -        return;
> > > > +        if ( handle_mmio(&info) )
> > > > +        {
> > > > +            advance_pc(regs, hsr);
> > > > +            return;
> > > > +        }
> > > > +        break;
> > > > +    default:
> > > > +        gprintk(XENLOG_WARNING, "Unsupported DFSC: HSR=%#x DFSC=%#x\n",
> > > > +                hsr.bits, dabt.dfsc);
> > > Given that bad_data_abort, which is right after, will print HSR again, I
> > > would remove it from this message as it's redundant.
> > I agree this is redundant, however the message below will only be printed in
> > debug build because a guest could trigger it easily. The message above is here
> > to catch any possible misconfiguration of stage-2 page table or hardware issue
> > (ECC error, address size, TLB conflict...) and could be seen in non-debug
> > build.
> Fair enough, it will just look a bit weird in the source code.
Would a comment in the source code make it less weird?

> Reviewed-by: Stefano Stabellini <sstabellini@xxxxxxxxxx>
--
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel