Re: [Qemu-devel] [PATCH v9 30/37] qapi: Canonicalize missing object to :empty

[Markus Armbruster <armbru@redhat.com>]

Markus Armbruster <armbru@redhat.com> writes:

> Eric Blake <eblake@redhat.com> writes:
>
>> Now that we elide unnecessary visits of empty types, we can
>> start using the special ':empty' type in more places.  By using
>> the empty type as the base class of every explicit struct or
>> union, and as the default data for any command or event, we can
>> simplify later logic in qapi-{visit,commands,event} by merely
>> checking whether the type is empty, without also having to worry
>> whether a type was even supplied.
>>
>> Note that gen_object() in qapi-types still has to check for a
>> base, because it is also called for alternates (which have no
>> base).
>
> What about the one in gen_visit_struct()?
>
>         if (base and not base.is_empty()) or members:
>             ret += mcgen('''
>         visit_type_%(c_name)s_fields(v, obj, &err);
>     ''',
>                          c_name=c_name(name))
>
>> No change to generated code.
>>
>> Signed-off-by: Eric Blake <eblake@redhat.com>
>>
>> ---
>> v9: squash in more related changes
>> v8: rebase to earlier changes
>> v7: rebase to earlier changes
>> v6: new patch
>> ---
>>  scripts/qapi-commands.py                | 17 +++++++------
>>  scripts/qapi-event.py                   |  5 ++--
>>  scripts/qapi-types.py                   |  4 +--
>>  scripts/qapi-visit.py                   | 12 +++++----
>>  scripts/qapi.py                         | 25 +++++++++---------
>>  tests/qapi-schema/event-case.out        |  2 +-
>>  tests/qapi-schema/flat-union-empty.out  |  1 +
>>  tests/qapi-schema/ident-with-escape.out |  1 +
>>  tests/qapi-schema/indented-expr.out     |  4 +--
>>  tests/qapi-schema/qapi-schema-test.out  | 45 ++++++++++++++++++++++++++++++---
>>  tests/qapi-schema/union-clash-data.out  |  2 ++
>>  tests/qapi-schema/union-empty.out       |  1 +
>>  12 files changed, 83 insertions(+), 36 deletions(-)

Missing: update to qapi-introspect.py.  At least the expressions like

    arg_type or self._schema.the_empty_object_type

need updating.

I got stuck in reviewing PATCH 31.  The result is probably fine, but the
patch itself is impenetrable for me.  I'm playing with the code to see
whether I can turn that into a constructive remark.