Re: [PATCH 2/5] migration: Use QAPI_CLONE_MEMBERS in migrate_params_apply

[Fabiano Rosas <farosas@suse.de>]

Prasad Pandit <ppandit@redhat.com> writes:

> On Thu, 22 Jan 2026 at 02:35, Fabiano Rosas <farosas@suse.de> wrote:
>> > On Wed, 14 Jan 2026 at 18:55, Fabiano Rosas <farosas@suse.de> wrote:
>> >>  void qmp_migrate_set_parameters(MigrationParameters *params, Error **errp)
>> >> @@ -1539,7 +1429,7 @@ void qmp_migrate_set_parameters(MigrationParameters *params, Error **errp)
>> >>              migrate_get_current()->has_block_bitmap_mapping = true;
>> >>          }
>> >>
>> >> -        migrate_params_apply(params);
>> >> +        migrate_params_apply(&tmp);
>> >
>> > * This change looks unrelated to the rest of this patch. I see it is
>> > done in another [PATCH 5/5] ... QAPI_MERGE patch, we should move this
>> > there.
>> >
>>
>> This is the only reason we can use QAPI_CLONE_MEMBERS. The params
>> contain only the information that came from QAPI. What we really want to
>> apply is what's in tmp. I refer to it in the commit message:
>>
>> "use the temporary object, which already contains the current migration
>> parameters plus the new ones and was just validated by
>> migration_params_check"
>
> * Yes, here we are changing from params to &tpm. And the other [PATCH
> 5/5] changes the same from
>
>      void qmp_migrate_set_parameters(MigrationParameters *params, Error **errp)
>               ...
>      -        migrate_params_apply(&tmp);
>     +        migrate_params_apply(tmp);
>
> along with other related changes. I was thinking we change
> 'migrate_params_apply' in [PATCH 2/5] and 'qmp_migrate_set_parameters'
> in [PATCH 5/5].
>

No, because then this patch breaks entirely. This patch can only exist
if tmp is being used. Or maybe I'm not following...

> Thank you.
> ---
>   - Prasad