Re: [PATCH v10 7/7] hw/ppc: Add emulation of Genesi/bPlan Pegasos II

[David Gibson]

On Wed, Mar 24, 2021 at 12:21:02PM +0100, BALATON Zoltan wrote:
> On Wed, 24 Mar 2021, David Gibson wrote:
> > On Tue, Mar 23, 2021 at 02:01:27PM +0100, BALATON Zoltan wrote:
> > > On Tue, 23 Mar 2021, David Gibson wrote:
> > > > On Wed, Mar 17, 2021 at 02:17:51AM +0100, BALATON Zoltan wrote:
> > > > > Add new machine called pegasos2 emulating the Genesi/bPlan Pegasos II,
> > > > > a PowerPC board based on the Marvell MV64361 system controller and the
> > > > > VIA VT8231 integrated south bridge/superio chips. It can run Linux,
> > > > > AmigaOS and a wide range of MorphOS versions. Currently a firmware ROM
> > > > > image is needed to boot and only MorphOS has a video driver to produce
> > > > > graphics output. Linux could work too but distros that supported this
> > > > > machine don't include usual video drivers so those only run with
> > > > > serial console for now.
> > > > > 
> > > > > Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
> > > > > Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
> > > > > ---
> > > > >  MAINTAINERS                             |  10 ++
> > > > >  default-configs/devices/ppc-softmmu.mak |   2 +
> > > > >  hw/ppc/Kconfig                          |   9 ++
> > > > >  hw/ppc/meson.build                      |   2 +
> > > > >  hw/ppc/pegasos2.c                       | 144 
> > > > > ++++++++++++++++++++++++
> > > > >  5 files changed, 167 insertions(+)
> > > > >  create mode 100644 hw/ppc/pegasos2.c
> > > > > 
> > > > > diff --git a/MAINTAINERS b/MAINTAINERS
> > > > > index b6ab3d25a7..1c3c55ef09 100644
> > > > > --- a/MAINTAINERS
> > > > > +++ b/MAINTAINERS
> > > > > @@ -1353,6 +1353,16 @@ F: pc-bios/canyonlands.dt[sb]
> > > > >  F: pc-bios/u-boot-sam460ex-20100605.bin
> > > > >  F: roms/u-boot-sam460ex
> > > > > 
> > > > > +pegasos2
> > > > > +M: BALATON Zoltan <balaton@eik.bme.hu>
> > > > > +R: David Gibson <david@gibson.dropbear.id.au>
> > > > > +L: qemu-ppc@nongnu.org
> > > > > +S: Maintained
> > > > > +F: hw/ppc/pegasos2.c
> > > > > +F: hw/pci-host/mv64361.c
> > > > > +F: hw/pci-host/mv643xx.h
> > > > > +F: include/hw/pci-host/mv64361.h
> > > > 
> > > > Oh, sorry about the comment in the previous patch.
> > > > 
> > > > >  RISC-V Machines
> > > > >  ---------------
> > > > >  OpenTitan
> > > > > diff --git a/default-configs/devices/ppc-softmmu.mak 
> > > > > b/default-configs/devices/ppc-softmmu.mak
> > > > > index 61b78b844d..4535993d8d 100644
> > > > > --- a/default-configs/devices/ppc-softmmu.mak
> > > > > +++ b/default-configs/devices/ppc-softmmu.mak
> > > > > @@ -14,5 +14,7 @@ CONFIG_SAM460EX=y
> > > > >  CONFIG_MAC_OLDWORLD=y
> > > > >  CONFIG_MAC_NEWWORLD=y
> > > > > 
> > > > > +CONFIG_PEGASOS2=y
> > > > 
> > > > I don't think we can have this default to enabled while it requires a
> > > > non-free ROM to start.
> > > 
> > > Not having it enabled though does not help those who might want to use it 
> > > as
> > > they are not people who can compile their own QEMU but rely on binaries so
> > > adding it without also enabling it is like it wasn't there at all in
> > > practice.
> > 
> > Not convinced, sorry.  If it's not usable out of the box, having to
> > build from source is kind of expected.  Or you could convince someone
> 
> I accept your point however there's a difference of only needing a ROM image
> to be able to use it from your distro's binary and having to rebuild it from
> source. So to me needing a ROM does not make it expected having to rebuild
> it. Needing to configure it some way would make that expected.

Well, you could always lobby the distro to include it.  Honestly I
think the distros would probably choose not to include something for
such a niche case, even if it was enabled by default upstream.   At
least.. if they were paying attention, which they're likely not.  But
sneaking something in via a side route is not something I'm going to
encourage.

> > (or do it yourself) to provide prebuild binaries for this purpose that
> > have the right things enabled.
> 
> There are people who provide binaries with patches for such purposes but
> that limits the availability of it compared to having it in all distro
> binaries without further effort. As I said I also plan to solve this
> eventually but I'd probably need VOF for that. Will that be merged at last?

I hope so, yes.  I've finally had some time to review recent rounds.
It'll be 6.1 material as well, obviously.  Note that there are some
issues you'll have to sort out to use this for Pegasos which doesn't
have an actual hypervisor mode.

> Other alternatives would be modifying SLOF, OpenBIOS or OpenFirmware or
> rewrite SmartFirmware to free it from its non-distributable parts but I
> think VOF would be a simpler way also avoiding adding another full OF
> implementation to QEMU that already has more than there should be.
> 
> Regards,
> BALATON Zoltan
> 
> > > I can attempt to make some guests boot without a ROM but since
> > > guests expect an OpenFirmware client interface, I'll need something to
> > > provide that. I'm waiting for VOF to be merged for this.
> > > 
> > > Regards,
> > > BALATON Zoltan
> > 
> > 
> > 


-- 
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