Re: [Qemu-devel] [PATCH -V2] kvm: Add a new machine property kvm_type

[Alexander Graf]

On 18.12.2013, at 18:51, Aneesh Kumar K.V <address@hidden> wrote:

> Alexander Graf <address@hidden> writes:
> 
>> On 07.11.2013, at 15:27, Aneesh Kumar K.V <address@hidden> wrote:
>> 
>>> From: "Aneesh Kumar K.V" <address@hidden>
>>> diff --git a/hw/ppc/e500plat.c b/hw/ppc/e500plat.c
>>> index 2e964b2..4be4b24 100644
>>> --- a/hw/ppc/e500plat.c
>>> +++ b/hw/ppc/e500plat.c
>>> @@ -17,6 +17,7 @@
>>> #include "hw/pci/pci.h"
>>> #include "hw/ppc/openpic.h"
>>> #include "kvm_ppc.h"
>>> +#include "kvmtype.h"
>>> 
>>> static void e500plat_fixup_devtree(PPCE500Params *params, void *fdt)
>>> {
>>> @@ -51,6 +52,7 @@ static QEMUMachine e500plat_machine = {
>>>    .desc = "generic paravirt e500 platform",
>>>    .init = e500plat_init,
>>>    .max_cpus = 32,
>>> +    .kvm_type = pr_kvm_type,
>> 
>> What about mpc8544ds? Also, e500plat can definitely support HV
>> (e500mc+) and PR (e500v2) mode. So you probably want the same check
>> here as you have in spapr.
> 
> I don't have access to the hardware to validate. So at this point, I
> would suggest we change the code such that we don't change the
> functionality.

Works for me. Please accept neither PR nor HV then, but only leave the default.

> 
>> 
>> Unfortunately the patch doesn't apply anymore. Would you care to
>> rebase it? I think it's also more targeted towards the kvm queue
>> rather than the ppc queue, so I'd either like an ack from Paolo or
>> would defer to his tree for inclusion :).
> 
> 
> Will resend

Thanks!


Alex