Re: [Qemu-devel] [PATCH v2 00/13] qapi: Test coverage & clean up generated code

[Luiz Capitulino <lcapitulino@redhat.com>]

On Thu, 13 Feb 2014 08:41:48 -0500
Luiz Capitulino <lcapitulino@redhat.com> wrote:

> On Thu, 13 Feb 2014 10:31:17 +0100
> Paolo Bonzini <pbonzini@redhat.com> wrote:
> 
> > Il 10/02/2014 15:20, Markus Armbruster ha scritto:
> > > Coverity is unhappy with the generated code.  Nothing serious, just
> > > heaps of valid DEADCODE defects topped off with a few bogus
> > > FORWARD_NULL defects.
> > >
> > > I had a look at the generator, and decided I don't want to mess with
> > > it without decent test coverage.  Unfortunately, a few features have
> > > been added without tests.  My first seven patches make the tests catch
> > > up.  tests/qapi-schema/qapi-schema-test.json now covers all mcgen() in
> > > scripts/qapi*.py, except for a few in qapi-commands.py that are
> > > conditional on -m.
> > >
> > > The next four patches clean up the generated code.
> > >
> > > I also reviewed null checks in visitor implementations.  The final two
> > > patches address the issues I found there.
> > >
> > > v2:
> > > * Previously separate patch "qapi: Fix licensing of scripts" revised
> > >   and included as "[PATCH v2 08/13] qapi: Fix licensing of scripts"
> > > * Style cleanups
> > > * New PATCH 12/13 and 13/13
> > >
> > > Markus Armbruster (13):
> > >   tests/qapi-schema: Actually check successful QMP command response
> > >   tests/qapi-schema: Cover optional command arguments
> > >   tests/qapi-schema: Cover simple argument types
> > >   tests/qapi-schema: Cover anonymous union types
> > >   tests/qapi-schema: Cover complex types with base
> > >   tests/qapi-schema: Cover union types with base
> > >   tests/qapi-schema: Cover flat union types
> > >   qapi: Fix licensing of scripts
> > >   qapi: Drop nonsensical header guard in generated qapi-visit.c
> > >   qapi: Drop unused code in qapi-commands.py
> > >   qapi: Clean up null checking in generated visitors
> > >   qapi: Clean up superfluous null check in qapi_dealloc_type_str()
> > >   qapi: Add missing null check to opts_start_struct()
> > >
> > >  qapi/opts-visitor.c                     |  4 +-
> > >  qapi/qapi-dealloc-visitor.c             |  4 +-
> > >  scripts/qapi-commands.py                | 24 +---------
> > >  scripts/qapi-types.py                   |  4 +-
> > >  scripts/qapi-visit.py                   | 20 ++++-----
> > >  scripts/qapi.py                         |  4 +-
> > >  tests/qapi-schema/qapi-schema-test.json | 24 +++++++++-
> > >  tests/qapi-schema/qapi-schema-test.out  | 19 +++++---
> > >  tests/test-qmp-commands.c               | 78 +++++++++++++++++++++++++++------
> > >  tests/test-qmp-input-strict.c           | 69 ++++++++++++++++++++++++++++-
> > >  tests/test-qmp-input-visitor.c          | 45 +++++++++++++++++--
> > >  tests/test-qmp-output-visitor.c         | 67 ++++++++++++++++++++++++++--
> > >  tests/test-visitor-serialization.c      | 14 +++---
> > >  13 files changed, 299 insertions(+), 77 deletions(-)
> > >
> > 
> > Thanks Markus!
> > 
> > Luiz, should this go in through your tree?
> 
> It could. Markus, what's your plan for this series? Is anyone going to
> pick it up?

Turns out it conflicts with something on my tree (probably Wenchao's series).

Can you please rebase on top of it Markus?