Re: [Qemu-devel] [PATCH v8 00/26] qapi: QMP introspection

[Daniel P. Berrange]

On Wed, Sep 16, 2015 at 01:06:03PM +0200, Markus Armbruster wrote:
> Series needs my "[PATCH for-2.4 0/2] qom: Fix misuse of Error API" to
> pass tests.  It's in Andreas's qom-next (currently at commit de1da05),
> so I based on that.
> 
> You can fetch it from my git://repo.or.cz/qemu/armbru.git branch
> qapi-introspect.
> 
> Review of v6 found a small flaw in the PATCH 20.  Fixing it will
> require a (possibly time-consuming) dive down the visitors rabbit
> hole.  To avoid further delay, we decided to go along with the flawed
> patch, then fix it up in a follow-up.  See PATCH 20 for details.
> 
> * PATCH 01-18 basically replace the intermediate representation.  The
>   replacement isn't complete, but fully functional.  See PATCH 02 for
>   rationale and future work.
> 
> * PATCH 19-24 replace the '**' type bypass by a proper top type called
>   'any'.
> 
> * PATCH 25-26 are the QMP introspection.
> 
> v8:
> * Non-trivial rebase due to commit 351d36e "qapi: allow override of
>   default enum prefix naming", R-bys dropped from the patches
>   mentioned below unless noted otherwise.
> * PATCH 02:
>   - Add prefix to QAPISchemaObjectType
> * PATCH 03:
>   - Use prefix in QAPISchemaObjectType.c_null()
> * PATCH 04:
>   - Add prefix to QAPISchemaVisitor.visit_enum_type()
> * PATCH 05:
>   - Use prefix in QAPISchemaTestVisitor.visit_enum_type()
> * PATCH 06:
>   - Use prefix in QAPISchemaGenTypeVisitor.visit_enum_type()
> * PATCH 07:
>   - Ignore prefix in QAPISchemaGenTypeVisitor.visit_enum_type()
> * PATCH 09:
>   - The code moved changed in commit 351d36e, redo code motion
> * PATCH 13:
>   - Resolve conflicts
> * PATCH 20:
>   - Commit message tweaked [Dan], R-bys retained
> * PATCH 21:
>   - Resolve expected test output conflicts by regenerating it, R-bys
>     retained
> * PATCH 25:
>   - Resolve straightforward conflicts in Makefile.objs, tests/Makefile
>   - Use prefix in QAPISchemaGenIntrospectVisitor.visit_enum_type()
>   - Resolve expected test output conflicts by regenerating it
> * Since the non-trivial rebase invalidated my testing, I put back the
>   scaffolding patches dropped in v6, and verified everything's still
>   sane.  I tagged this version qapi-introspect-v8-pre in my git
>   repository.
> * Unless this fails review, I will require further conflicting QAPI
>   changes to base on it.

I've compared a diff-of-diffs for v7 vs v8 and confirmed the conflict
resolution you did for my enum prefix change looks sane in all cases.
So I've replied with R-B on all the patches, which I'd previously
reviewed - that's all except 25 which I've never reviewed yet.

Regards,
Daniel
-- 
|: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org              -o-             http://virt-manager.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|