On Sat, Feb 26, 2011 at 4:12 PM, Juan Quintela <quint...@redhat.com> wrote:
> Blue Swirl <blauwir...@gmail.com> wrote:
>> On Thu, Feb 24, 2011 at 7:57 PM, Juan Quintela <quint...@redhat.com> wrote:
>>> +static const VMStateDescription vmstate_stellaris_button = {
>>> + .name = "stellaris_button",
>>> + .version_id = 1,
>>> + .minimum_version_id = 1,
>>> + .minimum_version_id_old = 1,
>>
>> For consistency with other structures, these version fields should be 0.
>
> From the part that you didn't posted.
>
> @@ -86,6 +85,5 @@ void stellaris_gamepad_init(int n, qemu_irq *irq, const int
> *keycode)
> }
> s->num_buttons = n;
> qemu_add_kbd_event_handler(stellaris_gamepad_put_key, s);
> - register_savevm(NULL, "stellaris_gamepad", -1, 1,
> - stellaris_gamepad_save, stellaris_gamepad_load, s);
> + vmstate_register(NULL, -1, &vmstate_stellaris_gamepad, s);
> }
> --
> 1.7.4
>
>
> Version is already at ' (4th argument of register_savevm). I just used
> the version that was there. I agree that some devices had started at
> version 0 and others at version 1, but that is independent of this series.
The patch introduces two VMState structures, main VMState for the
whole gamepad and a secondary VMState for button structures. The
version for main VMState is OK at 1, but I was talking about the
button substructure VMState version, which is 0 in other patches.
