Re: [PATCH] qapi: enable use of g_autoptr with QAPI types

[Daniel P . Berrangé]

On Thu, Jul 23, 2020 at 02:50:51PM +0200, Markus Armbruster wrote:
> Daniel P. Berrangé <berrange@redhat.com> writes:
> 
> > On Thu, Jul 23, 2020 at 06:49:44AM -0500, Eric Blake wrote:
> >> On 7/23/20 6:12 AM, Daniel P. Berrangé wrote:
> >> > Currently QAPI generates a type and function for free'ing it:
> >> > 
> >> >    typedef struct QCryptoBlockCreateOptions QCryptoBlockCreateOptions;
> >> >    void qapi_free_QCryptoBlockCreateOptions(QCryptoBlockCreateOptions 
> >> > *obj);
> >> > 
> >> 
> >> > The above code example now becomes
> >> > 
> >> >    g_autoptr(QCryptoBlockCreateOptions) opts = NULL;
> >> > 
> >> >    opts = g_new0(QCryptoBlockCreateOptions, 1);
> >> > 
> >> >    ....do stuff with opts...
> >> > 
> >> > Note, if the local pointer needs to live beyond the scope holding the
> >> > variable, then g_steal_pointer can be used. This is useful to return the
> >> > pointer to the caller in the success codepath, while letting it be freed
> >> > in all error codepaths.
> >> > 
> >> >    return g_steal_pointer(&opts);
> >> > 
> >> 
> >> Yep, the idea makes sense!
> 
> Agree.
> 
> >> > Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
> >> > ---
> >> >   include/crypto/block.h | 2 --
> >> >   scripts/qapi/types.py  | 1 +
> >> >   2 files changed, 1 insertion(+), 2 deletions(-)
> >> 
> >> Missing a counterpart change to docs/devel/qapi-code-gen.txt.
> 
> Yes.  Can do that in my tree.
> 
> >>                                                                And it might
> >> be nice to make this a series with at least one followup patch using the 
> >> new
> >> capability, or at least so 'make check' coverage.  But otherwise on the
> >> right track.
> >
> > The crypto/block.c already makes use of this capability, which is why
> > I had to remove the line from block.h to avoid declaring the same thing
> > twice !
> 
> Could be mentioned in the commit message.
> 
> Still, using it somewhere in tests would be nice.
> test-qobject-input-visitor.c's test_visitor_in_struct_nested() looks
> trivial to convert.  Feel free to pick something else.

Ok, I'll convert some.

Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|