Re: [Qemu-devel] [RFC PATCH 1/1] s390x/css: unresrict cssids

[Christian Borntraeger]

On 11/28/2017 01:14 PM, Cornelia Huck wrote:
> On Tue, 28 Nov 2017 12:49:04 +0100
> Boris Fiuczynski <address@hidden> wrote:
> 
>> On 11/28/2017 11:22 AM, Cornelia Huck wrote:
>>> On Tue, 28 Nov 2017 09:53:15 +0100
>>> Boris Fiuczynski <address@hidden> wrote:
>>>   
>>>> On 11/27/2017 05:56 PM, Cornelia Huck wrote:  
>>>>> Proposal 2: Export the default cssid as a machine property. If this
>>>>> property exists, it also implies that devices can be put into any css
>>>>> image (although it makes the most sense to put them into the default
>>>>> css image as indicated by the property). Can be made r/w later if it is
>>>>> too much for 2.12.  
>>>> Just as a side discussion:
>>>> I know of qom command query-machines but that does not seem to provide
>>>> the information you suggest to use with proposal 2.  
>> Sorry, I meant query-command-line-options.
>>
>>>> What qom command do you suggest to use for the introspection of the
>>>> machine options access mode?
>>>>  
>>>
>>> Is qom-get what you are looking for?
>>>
>>> virsh # qemu-monitor-command vm1 --pretty '{ "execute": "qom-get", 
>>> "arguments": { "path": "/machine/", "property": "accel"} }'
>>> {
>>>    "return": "kvm",
>>>    "id": "libvirt-18"
>>> }
>>>   
>> How do you find out from returned values that accel is r/o or r/w?
>>
> 
> I read some code and it turns out that we aren't really prepared for
> any r/o opts... sigh.
> 
> So, proposal 2 is only viable if we make it configurable from the start.

I really really dislike option 2.
Binding the capability to "assign freely" to a property that is named default 
css is
just wrong. Both capabilities are really independent.

I strongly prefer option 3.

> 
> Halil, do you see any chance to do this for 2.12? There's plenty of
> time left, and I don't think it's too hard. If not, we don't have any
> other option than proposal 3, even though I don't like it a lot.
>