Re: [Qemu-devel] [PATCH v2 3/9] s390x: remove s390-virtio devices

[Pierre Morel]

On 01/11/2016 07:45 PM, Cornelia Huck wrote:
> On Mon, 11 Jan 2016 18:12:27 +0100
> Thomas Huth <address@hidden> wrote:
>
> >   Happy new year!
> To you as well :)
>
> > On 07.01.2016 14:36, Cornelia Huck wrote:
> > > From: Pierre Morel <address@hidden>
> > >
> > > The s390-virtio machine has been removed; remove the associated devices
> > > as well.
> > >
> > > hw/s390x/s390-virtio-bus.c and hw/s390x/s390-virtio-bus.h
> > > have been deleted and removed from hw/s390x/Makefile.objs
> > >
> > > virtio-size has no more meaning for the modern machine
> > > and has been removed from helper.c and cpu.h
> > >
> > > virtio-serial-s390 belonging to the old machine is
> > > being removed from vl.c
> > >
> > > Signed-off-by: Pierre Morel <address@hidden>
> > > Acked-by: Cornelia Huck <address@hidden>
> > > Signed-off-by: Cornelia Huck <address@hidden>
> > > ---
> > >   hw/s390x/Makefile.objs     |   2 +-
> > >   hw/s390x/s390-virtio-bus.c | 758 ---------------------------------------------
> > >   hw/s390x/s390-virtio-bus.h | 186 -----------
> > >   hw/s390x/s390-virtio.c     |   1 -
> > >   target-s390x/cpu.h         |   3 -
> > >   target-s390x/helper.c      |   2 +-
> > >   vl.c                       |   7 +-
> > >   7 files changed, 3 insertions(+), 956 deletions(-)
> > >   delete mode 100644 hw/s390x/s390-virtio-bus.c
> > >   delete mode 100644 hw/s390x/s390-virtio-bus.h
> > [...]
> > > diff --git a/target-s390x/helper.c b/target-s390x/helper.c
> > > index aa58f39..e23fdeb 100644
> > > --- a/target-s390x/helper.c
> > > +++ b/target-s390x/helper.c
> > > @@ -133,7 +133,7 @@ int s390_cpu_handle_mmu_fault(CPUState *cs, vaddr 
> > > orig_vaddr,
> > >       }
> > >   
> > >       /* check out of RAM access */
> > > -    if (raddr > (ram_size + virtio_size)) {
> > > +    if (raddr > (ram_size)) {
> > While you're at it, you could remove the parentheses around "ram_size"
> > here ... I know, you like parentheses, Cornelia ... but here they IMHO
> > look quite superfluous now...
> That's more an oversight; will remove.
>
> > >           DPRINTF("%s: raddr %" PRIx64 " > ram_size %" PRIx64 "\n", __func__,
> > >                   (uint64_t)raddr, (uint64_t)ram_size);
> > >           trigger_pgm_exception(env, PGM_ADDRESSING, ILEN_LATER);
> > > diff --git a/vl.c b/vl.c
> > > index 5aaea77..0c69d94 100644
> > > --- a/vl.c
> > > +++ b/vl.c
> > > @@ -227,7 +227,6 @@ static struct {
> > >       { .driver = "ide-drive",            .flag = &default_cdrom     },
> > >       { .driver = "scsi-cd",              .flag = &default_cdrom     },
> > >       { .driver = "virtio-serial-pci",    .flag = &default_virtcon   },
> > > -    { .driver = "virtio-serial-s390",   .flag = &default_virtcon   },
> > >       { .driver = "virtio-serial",        .flag = &default_virtcon   },
> > >       { .driver = "VGA",                  .flag = &default_vga       },
> > >       { .driver = "isa-vga",              .flag = &default_vga       },
> > > @@ -2548,11 +2547,7 @@ static int virtcon_parse(const char *devname)
> > >       }
> > >   
> > >       bus_opts = qemu_opts_create(device, NULL, 0, &error_abort);
> > > -    if (arch_type == QEMU_ARCH_S390X) {
> > > -        qemu_opt_set(bus_opts, "driver", "virtio-serial-s390", &error_abort);
> > > -    } else {
> > > -        qemu_opt_set(bus_opts, "driver", "virtio-serial-pci", &error_abort);
> > > -    }
> > > +    qemu_opt_set(bus_opts, "driver", "virtio-serial", &error_abort);
> > Are you sure you want to change the default value for non-s390 systems
> > from "virtio-serial-pci" to "virtio-serial", too?
> > Also, isn't the virtio-ccw variant called "virtio-serial-ccw" instead or
> > did I miss something?
> That's OK, virtio-serial will point to the right driver. And I trust
> Pierre to have verified it :)

I verified it only on the x86_64 architecture, lack of hardware.
I think that if it is not broken there, there should be no problem for 
others,
but of course let me know if you find any issue.