Re: [Qemu-devel] [PATCH v4 04/16] pc: forbid BSP removal

[Bandan Das <bsd@redhat.com>]

Eduardo Habkost <ehabkost@redhat.com> writes:

> On Thu, Jul 14, 2016 at 06:54:33PM +0200, Igor Mammedov wrote:
>> BSP is assumed to always present in QEMU code, so
>> untile that assumptions are gone, deny removal request.
>> In another words QEMU won't support BSP hot-unplug.
>> 
>> Signed-off-by: Igor Mammedov <imammedo@redhat.com>
>> ---
>>  hw/i386/pc.c | 7 +++++++
>>  1 file changed, 7 insertions(+)
>> 
>> diff --git a/hw/i386/pc.c b/hw/i386/pc.c
>> index 5a67f15..33c5f97 100644
>> --- a/hw/i386/pc.c
>> +++ b/hw/i386/pc.c
>> @@ -1751,10 +1751,17 @@ out:
>>  static void pc_cpu_unplug_request_cb(HotplugHandler *hotplug_dev,
>>                                       DeviceState *dev, Error **errp)
>>  {
>> +    int idx;
>>      HotplugHandlerClass *hhc;
>>      Error *local_err = NULL;
>>      PCMachineState *pcms = PC_MACHINE(hotplug_dev);
>>  
>> +    pc_find_cpu_slot(pcms, CPU(dev), &idx);
>
> Looks fragile: if one day we create any TYPE_CPU object that is
> not in possible_cpus array, idx is undefined. I suggest
> initializing idx to -1 above.

Or just let pc_find_cpu_slot universally set it to -1 since
this series assumes that -1 means index isn't valid.

> I can change it when committing, if that's OK for you.
>
>
>> +    if (idx == 0) {
>> +        error_setg(&local_err, "1st CPU (BSP) is unpluggable");
>> +        goto out;
>> +    }
>> +
>>      hhc = HOTPLUG_HANDLER_GET_CLASS(pcms->acpi_dev);
>>      hhc->unplug_request(HOTPLUG_HANDLER(pcms->acpi_dev), dev, &local_err);
>>  
>> -- 
>> 2.7.4
>>