Re: [Qemu-ppc] [Qemu-devel] [PATCH 7/7] prep: QOM'ify System I/O

[Hervé Poussineau]

Andreas Färber a écrit :
> Am 02.05.2013 22:09, schrieb Hervé Poussineau:
> > Most of the functionality is extracted from hw/ppc/prep.c.
> > Also add support for board identification/equipment registers.
> >
> > Document it for the IBM 43p emulation.
> >
> > Cc: Julio Guerra <address@hidden>
> > Signed-off-by: Hervé Poussineau <address@hidden>
> > ---
> >  docs/ibm_43p.cfg       |    5 +
> >  hw/ppc/Makefile.objs   |    1 +
> >  hw/ppc/prep_systemio.c |  298 ++++++++++++++++++++++++++++++++++++++++++++++++
> >  trace-events           |    4 +
> >  4 files changed, 308 insertions(+)
> >  create mode 100644 hw/ppc/prep_systemio.c
>
> Haven't reviewed the full patch yet, but since this is not modifying
> hw/ppc/prep.c, it is duplicating code rather than QOM'ifying the
> existing code.
>
> Have you looked into Julio's patch whom you CC? I'm still not sure how
> to solve things for 1.5 (and this series a consider -next).

Yes, I've partly taken Julio's patch into account, ie port 0x92 is now 
read/write. However, I didn't change the way the reset is done due to 
missing agrement of how it should be done: "this is touching on the same 
soft reset topic that I am awaiting the outcome for x86". [1]

Moreover, registers emulated are not exactly the same as in hw/ppc/prep.c:
Registers not present in hw/ppc/prep_systemio.c:
0x800: Motorola CPU configuration register
0x802: Motorola base module feature register
0x803: Motorola base module status register
0x823: Something related to no L2 cache?
Those seem specific to Motorola, so they probably belong to another device.

New registers added to hw/ppc/prep_systemio.c:
0x818: Key lock (Read Only)
0x852: System Board Identification (Read Only)

I'm not really sure of impacts of changing 'prep' machine to remove some 
registers and adding new ones, so I prefered to keep it as is. However, 
if you think that I should use the QOM'ified System I/O device in 'prep' 
machine (and this will change emulated machine), I'll do it.

Hervé

[1] https://lists.gnu.org/archive/html/qemu-devel/2013-04/msg03359.html