On Tue, Sep 03, 2013 at 02:32:52PM +0200, Paolo Bonzini wrote:
> This becomes important when undoing realize's initializations is split
> in two places (unrealize and exit).
> 
> The way to fix this could be to split realize further into "alloc" (done
> once) and "realize" (can be undone).

Or alloc + init ?

> 
> Signed-off-by: Paolo Bonzini <pbonz...@redhat.com>
> ---
>  include/hw/qdev-core.h | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/include/hw/qdev-core.h b/include/hw/qdev-core.h
> index 46972f4..d840f06 100644
> --- a/include/hw/qdev-core.h
> +++ b/include/hw/qdev-core.h
> @@ -86,6 +86,10 @@ struct VMStateDescription;
>   * object_initialize() in their own #TypeInfo.instance_init and forward the
>   * realization events appropriately.
>   *
> + * Note that for now it is not possible to unrealize a device and then
> + * re-realize it.  While this may change in the future, it is a valid
> + * assumption for now.
> + *
>   * The @init callback is considered private to a particular bus 
> implementation
>   * (immediate abstract child types of TYPE_DEVICE). Derived leaf types set an
>   * "init" callback on their parent class instead.
> -- 
> 1.8.3.1
> 

Reply via email to