On 05/10/2013 03:19 PM, Martin Kletzander wrote:
On 04/30/2013 06:21 PM, Eric Blake wrote:
> On 04/30/2013 09:18 AM, John Ferlan wrote:
>> On 04/30/2013 10:42 AM, Martin Kletzander wrote:
>>> Adding support for new attribute 'websocket' in the
'<graphics>'
>>> element, the attribute value is the port to listen on with '-1'
>>> meaning auto-allocation, '0' meaning no websockets.
>>>
>>> Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com>
>>> ---
>
>>>
>>> + if (websocket) {
>>> + if (virStrToLong_i(websocket,
>>> + NULL, 10,
>>> + &def->data.vnc.websocket) < 0) {
>>> + virReportError(VIR_ERR_INTERNAL_ERROR,
>>> + _("cannot parse vnc websocket port
%s"), port);
>>
>> I think you meant websocket, not port for the string... it's a
>> cut-n-paste error it seems.
>>
>> "port" will be VIR_FREE()'d already (or it may not have been found,
so
>> it's already NULL.
>
> But I like the proposal, and agree with the docs calling it out as 1.0.6
> material.
>
> ACK with John's finding fixed.
>
I fixed what John found out and will push the series after I'll check
one more thing. Sorry it took so long, I just found out about this mail
since I wasn't one of the recipients.
I've found out there was missing call for virPortAllocatorRelease, plus
the number of comments made me send a v3 just to be sure. So rebased
version is available now [1].
Martin
[1]
http://www.redhat.com/archives/libvir-list/2013-May/msg00906.html