On Wed, Jan 10, 2018 at 02:33:41PM +0100, Paolo Bonzini wrote: > On 10/01/2018 02:55, David Gibson wrote: > > On Tue, Jan 09, 2018 at 03:15:25PM +0000, Peter Maydell wrote: > >> On 9 January 2018 at 12:16, David Gibson <da...@gibson.dropbear.id.au> > >> wrote: > >>> Thanks. Even if you can identify which patch it is and we can > >>> postpone that one would be a bug help. > >> > >> Bisection blames this one: > >> > >> pbonz...@redhat.com (1): > >> target-ppc: optimize cmp translation > > > > Aha! Ok, I'll pull that one out until we find a fix. > > Richard posted one, it's an ARM32 backend bug. > > > [snip] > >> (This is a build for arm32, running in a chroot on an aarch64 > >> box, non-debug build.) > > > > Ah.. but since it's a chroot, still an aarch64 cpu behind it all. > > No, it's running 32-bit ARM code.
Ah. But maybe a different version to older ARM32s?
I'm trying to figure out why I entirely failed to reproduce this on a
Raspberry Pi 1 (and a couple of other arm32 machines, but I forget the
details of them).
> I
> > think the machines I managed to borrow were all much older slower
> > really truly arm32 machines, which I guess didn't show the problem.
> >
> > Oh.. and probably a debug (default) build as well.
> >
>
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
signature.asc
Description: PGP signature
