Re: Variant tests failing on big endian machines

[Torsten Schoenfeld <kaffeetisch gmx de>]

On 26.06.2015 15:23, Rafael dos Santos wrote:
> > the variant tests added with commit d0c52b8bb are failing on big endian
> > machines. As it can be seen from the log output below, the VariantClass
> > field is always empty.
> >
> > Could somebody help me debug the issue and/or point me on the right direction
> > about looking into it?
>
> I found the culprit. On the GVariant.xs file, we have:
>
> const char * g_variant_classify (GVariant *value);
>     PREINIT:
>         GVariantClass vclass;
>     CODE:
>         vclass = g_variant_classify (value);
>         RETVAL = (const char *) &vclass;
>     OUTPUT:
>         RETVAL
>
> The RETVAL value will be arch-dependent because of the way it's accessed. I'm not familiar with .xs syntax, 
> so what's the way to deal with endianness on this type of files?

Does this work?

const char * g_variant_classify (GVariant *value);
    PREINIT:
        GVariantClass vclass;
    CODE:
        vclass = g_variant_classify (value);
        RETVAL = (const char *) &((char) vclass);
    OUTPUT:
        RETVAL