On Fri, Mar 11, 2011 at 09:56:42AM -0700, Eric Blake wrote:
On 03/11/2011 08:05 AM, Jiri Denemark wrote:
>> + if (qemuMonitorJSONHasError(reply, "CommandNotFound")) {
>> + VIR_DEBUG0(_("drive_del command not found, trying HMP"));
>> + ret = qemuMonitorTextDriveDel(mon, drivestr);
>> + } else if (qemuMonitorJSONHasError(reply, "DeviceNotFound")) {
>> + /* NB: device not found errors mean the drive was
>> + * auto-deleted and we ignore the error */
>> + ret = 0;
>> + } else {
>> + ret = qemuMonitorJSONCheckError(cmd, reply);
>> }
>
> Looks good, although I think we should issue the "deleting disk is not
> supported. This may leak data if disk is reassigned" error also in case
> human-monitor-command is not supported. But that will need some more work on
> the HMP infrastructure since it's not possible to get this from
> qemuMonitorText* function that we call.
Do we want to push this now, or get the infrastructure improvements in
first?
What about get it in first based on the fact that human-monitor-command
is supported? (with removing the two _() in VIR_DEBUG0)
--
Thanks,
Hu Tao