Re: [PATCH] tests/acceptance/virtio_check_params: Disable test

["Philippe Mathieu-Daudé" <philmd@redhat.com>]

On 2/6/20 6:24 PM, Cornelia Huck wrote:
> On Thu,  6 Feb 2020 18:17:15 +0100
> Philippe Mathieu-Daudé <philmd@redhat.com> wrote:
> 
>> This test fails on various CI:
>>
>> - Using QEMU 4.0:
>>
>>    tests/acceptance/x86_cpu_model_versions.py:X86CPUModelAliases.test_none_alias:  ERROR: 'alias-of' (0.45 s)
>>
>> - On OSX
>>
>>     Unexpected error in object_property_find() at qom/object.c:1201:
>>     qemu-system-x86_64: -device virtio-blk-pci,id=scsi0,drive=drive0: can't apply global virtio-blk-device.scsi=true: Property '.scsi' not found
> 
> Hm, but there was a fix for that, wasn't there?

Yes, the patch is on the list:
https://www.mail-archive.com/qemu-devel@nongnu.org/msg675259.html

Michael S. Tsirkin takes virtio, I'll ping him to see if he wants a 
respin or can take it and amend your suggestions.

>>
>> - When removing unavailable machine:
>>
>>    VirtioMaxSegSettingsCheck.test_machine_types: ERROR: list.remove(x): x not in list (0.12 s)
>>
>> - Using Xen:
>>
>>    xencall: error: Could not obtain handle on privileged command interface: No such file or directory
>>    xen be core: xen be core: can't open xen interface
>>
>> - On PPC:
>>
>>    TestFail: machine type pseries-2.8: No Transactional Memory support in TCG, try appending -machine cap-htm=off
>>
>> - On S390X:
>>
>>    ERROR: qemu-system-s390x: -device virtio-scsi-pci,id=scsi0: 'virtio-scsi-pci' is not a valid device model name
> 
> Depends on how your QEMU is configured... does virtio-pci default to
> off on s390x? I use it all the time to test pci :)

Yes, this is a --without-default-devices job indeed.

>>
>> Disable it for now.
>>
>> Cc: Denis Plotnikov <dplotnikov@virtuozzo.com>
>> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> ---
>> Series fixing the errors:
>> https://www.mail-archive.com/qemu-devel@nongnu.org/msg675074.html
>> ---
>>   tests/acceptance/virtio_check_params.py | 2 ++
>>   1 file changed, 2 insertions(+)
>>
> 
> It seems reasonable though to disable it until the problems are fixed.

The patches are on the list, they simply need review :)

> Acked-by: Cornelia Huck <cohuck@redhat.com>

Thanks,

Phil.