On 04/09/2015 10:09 AM, Peter Krempa wrote:
...
>>
>> Just checking...
>>
>> This change is essentially the same as in qemuDomainBlockPullCommon
>> where if (!modern) {} was added right?
>
> Yes. This one would be redundant.
>
>>
>>> -
>>> - if (backingName && mode != BLOCK_JOB_PULL) {
>>> - virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
>>> - _("backing name is supported only for block pull"));
>>> - return -1;
>>> - }
>>
>> And this won't be necessary.... since we no longer have multiple
>> (ab)users of the same API
>
> Exactly.
>
>>
>>> -
>>> - if (backingName && !base) {
>>> - virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
>>> - _("backing name requires a base image"));
>>> - return -1;
>>> - }
>>
>> Is there a check for this somewhere that I missed?
>
> The caller ensures that this does not happen. We could leave this one
> possibly in if you want.
>
Not a requirement (but at least you know I read the code ;-)). It's one
of those paranoia things.
John
>>
>>> + const char *cmd_name = modern ? "block-stream" : "block_stream";
>>>
>>> - if (speed && mode == BLOCK_JOB_PULL && !modern) {
>>> - virReportError(VIR_ERR_INTERNAL_ERROR,
>>> - _("only modern block pull supports speed: %llu"),
>>> - speed);
>>> - return -1;
>>> - }
>>
>> And this is the second half of the check in qemuDomainBlockPullCommon
>>
>> ACK - in general - Just want to make sure the "if (backingName && !base)
>> wasn't erroneously removed.
>>
>> John
>
> Peter
>
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list