[libvirt] [PATCH v2] Correct include-password option and rewrite domdisplay

Martin Kletzander mkletzan at redhat.com
Thu Nov 22 13:45:48 UTC 2012


On 11/22/2012 02:10 PM, Peter Krempa wrote:
> On 11/22/12 11:34, Martin Kletzander wrote:
>> The 'virsh domdisplay' command is able to display the password
>> configured for spice, but it was missing for vnc type graphics.
>> Also, there were some inconsistencies that are cleaned now.
>> ---
[...]
>> +        }
>> +
>>           /* Then host name or IP */
>>           if (!listen_addr || STREQ((const char *)listen_addr,
>> "0.0.0.0"))
>>               virBufferAddLit(&buf, "localhost");
>> @@ -7115,20 +7134,11 @@ cmdDomDisplay(vshControl *ctl, const vshCmd *cmd)
>>           VIR_FREE(listen_addr);
>>
>>           /* Add the port */
>> -        if (STREQ(scheme[iter], "spice"))
>> -            virBufferAsprintf(&buf, "?port=%d", port);
>> -        else
>> -            virBufferAsprintf(&buf, ":%d", port);
>> +        virBufferAsprintf(&buf, ":%d", port);
>>
>>           /* TLS Port */
>>           if (tls_port)
>> -            virBufferAsprintf(&buf, "&tls-port=%d", tls_port);
>> -
>> -        /* Password */
>> -        if (passwd) {
>> -            virBufferAsprintf(&buf, "&password=%s", passwd);
>> -            VIR_FREE(passwd);
>> -        }
>> +            virBufferAsprintf(&buf, "?tls-port=%d", tls_port);
>>
>>           /* Ensure we can print our URI */
>>           if (virBufferError(&buf)) {
>>
> 
> I'm not sure about the change of the password parameter. Could you back
> that up somehow?
> 

Unfortunately I cannot.  spicy is unable to parse the new version
correctly, but I believe that's a bug since there is a common knowledge
where to put the password.  I cooked up a win/win version with the spice
password being printed out the old way and vnc the new (standard) way,
so hopefully everyone could be satisfied, but that seems *very*
inconsistent to me.  But since I also shrunk the code a bit more and
fixed one more thing there, I'll send it and let's hope we'll come to
some conclusion then.

Martin




More information about the libvir-list mailing list