Daniel P. Berrange wrote:
> On Mon, Jun 02, 2014 at 01:53:46PM +0100, Ian Campbell wrote:
>
>> On Fri, 2014-05-30 at 18:24 +0100, Daniel P. Berrange wrote:
>>
>>> + if (STRNEQ(expectargv, (char *)actualargv)) {
>>> + virtTestDifference(stderr, expectargv, (char *)actualargv);
>>> + goto cleanup;
>>> + }
>>>
>> Since you are using libxl_domain_config_gen_json you can control the
>> pretty printing, but if you were to use the libxl_domain_config_to_json
>> you might have problems if the library was to do something slightly
>> different e.g. with whitespace.
>>
>> In 4.5 we will have libxl_*_from_json and (I think) libxl_*_compare, so
>> you could read in the template and compare it with the generated struct.
>> That doesn't help you now of course.
>>
>> Also in 4.5 the json will omit fields which are set to the their
>> explicit default value. libxl_*_from_json will still do the right thing,
>> but it'd be another annoyance for you here I think.
>>
>> Lastly, when we add new fields to the API they will start showing up in
>> the json (modulo the omission of defaults discussed above).
>>
>
> Hmm, that's a v good point.
>
One that can already be seen. The test works when run on Xen 4.3, but
fails on 4.4. E.g. c_info gained some new fields
"poolid": 0,
"run_hotplug_scripts": "<default>",
+ "pvh": "<default>",
+ "driver_domain": "<default>"
},
And on_watchdog's value changed
"on_poweroff": null,
"on_reboot": "destroy",
- "on_watchdog": null,
+ "on_watchdog": "destroy",
"on_crash": "destroy"
Regards,
Jim
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list