On 04/15/2015 12:13 PM, John Snow wrote:
>
>
> On 04/15/2015 11:19 AM, Eric Blake wrote:
>> Pretending that QMP doesn't understand a command merely because
>> we are not in the right mode doesn't help first-time users figure
>> out what to do to correct things. Although the documentation for
>> QMP calls out capabilities negotiation, we should also make it
>> clear in our error messages what we were expecting. With this
>> patch, I now get the following transcript:
>>
>> $ ./x86_64-softmmu/qemu-system-x86_64 -qmp stdio -nodefaults
>> {"QMP": {"version": {"qemu": {"micro": 93, "minor": 2, "major": 2},
>> "package": ""}, "capabilities": []}}
>> {"execute":"huh"}
>> {"error": {"class": "CommandNotFound", "desc": "The command huh has
>> not been found"}}
>> {"execute":"quit"}
>> {"error": {"class": "CommandNotFound", "desc": "Expecting capabilities
>> negotiation with 'qmp_capabilities' before command 'quit'"}}
>
> Any particular reason why we should keep the "CommandNotFound" error
> class here? Backwards compat? Inertia?
>
>> {"execute":"qmp_capabilities"}
>> {"return": {}}
>> {"execute":"qmp_capabilities"}
>> {"error": {"class": "CommandNotFound", "desc": "Capabilities
>> negotiation is already complete, command 'qmp_capabilities' ignored"}}
>
> Same here.Backwards compat. I can't prove that anyone else was relying on specific classes (in particular, although it is unlikely that anyone was issuing qmp_capabilities more than once, or cares what error class was returned, it IS a useful test for probing if the connection is in capability negotiation mode when reconnecting to a monitor after a libvirtd restart). It's better to be conservative and avoid changing the error class (which must be reliable to machine readers) and only impact the error message (which is human readable and is documented to not be machine parseable, so we can change that at will). -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org
signature.asc
Description: OpenPGP digital signature
