On 09/22/2016 01:48 PM, Marc-André Lureau wrote: > These 2 tests exhibit two qmp bugs fixed by the previous patches. > > Signed-off-by: Marc-André Lureau <marcandre.lur...@redhat.com> > Reviewed-by: Daniel P. Berrange <berra...@redhat.com> > --- > tests/test-qemu-qmp.c | 61 > ++++++++++++++++++++++++++++++++++++++++++++++++++ > tests/Makefile.include | 2 ++ > 2 files changed, 63 insertions(+) > create mode 100644 tests/test-qemu-qmp.c > > diff --git a/tests/test-qemu-qmp.c b/tests/test-qemu-qmp.c > new file mode 100644 > index 0000000..3cdee59 > --- /dev/null > +++ b/tests/test-qemu-qmp.c > @@ -0,0 +1,61 @@ > +#include "qemu/osdep.h" > +#include "libqtest.h" > +
No copyright blurb? The default GPLv2+ is fine, but being explicit never
hurts.
> +
> +static void test_object_add_without_props(void)
> +{
> + QDict *ret, *error;
> + const gchar *class, *desc;
> +
> + ret = qmp("{'execute': 'object-add',"
> + " 'arguments': { 'qom-type': 'memory-backend-ram', 'id':
> 'ram1' } }");
> + g_assert_nonnull(ret);
> +
> + error = qdict_get_qdict(ret, "error");
> + class = qdict_get_try_str(error, "class");
This variable won't compile with C++; should you rename it to 'klass'
just in case, since we already have part of qemu being built by C++ for
the sake of qga on Windows?
> + desc = qdict_get_try_str(error, "desc");
> +
> + g_assert_cmpstr(class, ==, "GenericError");
> + g_assert_cmpstr(desc, ==, "can't create backend with size 0");
A little bit fragile to be parsing for a particular error message
wording, but I guess I can live with it.
Reviewed-by: Eric Blake <ebl...@redhat.com>
--
Eric Blake eblake redhat com +1-919-301-3266
Libvirt virtualization library http://libvirt.org
signature.asc
Description: OpenPGP digital signature
