On Mon, Sep 21, 2015 at 07:22:51PM +0200, Thomas Huth wrote:
> On 21/09/15 18:39, Markus Armbruster wrote:
> > Thomas Huth <th...@redhat.com> writes:
> >> On 21/09/15 08:14, Markus Armbruster wrote:
> >>> Eduardo Habkost <ehabk...@redhat.com> writes:
> >>>> On Fri, Sep 18, 2015 at 08:42:54PM +0200, Thomas Huth wrote:
> >>>>> On 18/09/15 14:00, Markus Armbruster wrote:
[...]
> >>>>>> * Assert kvm_enabled(): "host-x86_64-cpu", host-i386-cpu",
> >>>>>> "host-powerpc64-cpu", "host-embedded-powerpc-cpu",
> >>>>>> "host-powerpc-cpu"
> >>>>>
> >>>>> I just had a look at the powerpc code - you're likely talking about
> >>>>> the "assert(kvm_enabled());" in the kvmppc_host_cpu_initfn() in
> >>>>> target-ppc/kvm.c ? That should be fine, I think, because
> >>>>> kvm_ppc_register_host_cpu_type() is only done on ppc when KVM has been
> >>>>> enabled.
> >>>
> >>> Easy to verify on a KVM-capable PPC host: try -device C,help with KVM on
> >>> and off, where C is the appropriate host CPU.
> >>
> >> In both cases (KVM on and off), I simply get:
> >>
> >> # qemu-system-ppc64 -machine accel=tcg -device host-powerpc64-cpu,help
> >> 'host-powerpc64-cpu' is not a valid device model name
> >> # qemu-system-ppc64 -machine accel=kvm -device host-powerpc64-cpu,help
> >> 'host-powerpc64-cpu' is not a valid device model name
> >>
> >> No crash/abort here.
> >
> > No introspection, either. Can you try the same with QMP command
> > device-list-properties?
>
> # ppc64-softmmu/qemu-system-ppc64 -nodefaults -S -display none -qmp stdio
[...]
> { "execute": "device-list-properties", "arguments": { "typename":
> "host-powerpc64-cpu" } }
> {"error": {"class": "DeviceNotFound", "desc": "Device 'host-powerpc64-cpu'
> not found"}}
>
> # ppc64-softmmu/qemu-system-ppc64 -machine accel=kvm -nodefaults -S -display
> none -qmp stdio
[...]
> { "execute": "device-list-properties", "arguments": { "typename":
> "host-powerpc64-cpu" } }
> {"return": [{"name": "compat", "description": "compatibility mode,
> power6/power7/power8", "type": "str"}]}
So, introspection is already broken, but inconsistently broken (because
it depends on machine arguments). Setting
cannot_even_create_with_object_new_yet will make it more consistently
broken.
--
Eduardo
