David Hildenbrand <da...@redhat.com> writes:
> On 22.08.2017 16:24, Cornelia Huck wrote:
>> On Tue, 22 Aug 2017 15:22:52 +0200
>> Marc-André Lureau <marcandre.lur...@redhat.com> wrote:
>>
>>> Signed-off-by: Marc-André Lureau <marcandre.lur...@redhat.com>
>>> ---
>>> qapi-schema.json | 10 +++++++---
>>> include/sysemu/arch_init.h | 6 ------
>>> monitor.c | 14 --------------
>>> qmp.c | 14 --------------
>>> stubs/arch-query-cpu-model-baseline.c | 12 ------------
>>> stubs/arch-query-cpu-model-comparison.c | 12 ------------
>>> target/s390x/cpu_models.c | 4 ++--
>>> stubs/Makefile.objs | 2 --
>>> 8 files changed, 9 insertions(+), 65 deletions(-)
>>> delete mode 100644 stubs/arch-query-cpu-model-baseline.c
>>> delete mode 100644 stubs/arch-query-cpu-model-comparison.c
>>>
>>> diff --git a/qapi-schema.json b/qapi-schema.json
>>> index 58574b3044..d4e1552ddc 100644
>>> --- a/qapi-schema.json
>>> +++ b/qapi-schema.json
>>> @@ -3577,7 +3577,8 @@
>>> #
>>> ##
>>> { 'command': 'dump-skeys',
>>> - 'data': { 'filename': 'str' } }
>>> + 'data': { 'filename': 'str' },
>>> + 'if': ['defined(NEED_CPU_H)', 'defined(TARGET_S390X)']}
>>
>> I agree with making this s390x specific...
>>>
>>> ##
>>> # @netdev_add:
>>> @@ -4621,7 +4622,9 @@
>>> ##
>>> { 'command': 'query-cpu-model-comparison',
>>> 'data': { 'modela': 'CpuModelInfo', 'modelb': 'CpuModelInfo' },
>>> - 'returns': 'CpuModelCompareInfo' }
>>> + 'returns': 'CpuModelCompareInfo',
>>> + 'if': ['defined(NEED_CPU_H)', 'defined(TARGET_S390X)']}
>>> +
>>>
>>> ##
>>> # @CpuModelBaselineInfo:
>>> @@ -4673,7 +4676,8 @@
>>> { 'command': 'query-cpu-model-baseline',
>>> 'data': { 'modela': 'CpuModelInfo',
>>> 'modelb': 'CpuModelInfo' },
>>> - 'returns': 'CpuModelBaselineInfo' }
>>> + 'returns': 'CpuModelBaselineInfo',
>>> + 'if': ['defined(NEED_CPU_H)', 'defined(TARGET_S390X)']}
>>
>> ...but I'm not sure about the cpu model stuff. Wasn't the idea to move
>> to this model for all architectures later? (Given that we have stubs
>> for architectures not implementing this, instead of ifdeffing it in
>> monitor.c)
>>
>
> +1, not architecture specific (in contrast to skey), simply not
> supported _yet_ on other architectures.
Yes, but Marc-André's patch makes the "not supported yet" information
available in query-qmp-schema. Carrying such information is pretty much
the point of schema introspection.
We could add a comment explaining this command isn't target-specific,
but just happens not to be implemented for some targets. Would that
help?
