* Sascha Silbe (si...@linux.vnet.ibm.com) wrote:
> Dear David,
>
> "Dr. David Alan Gilbert" <dgilb...@redhat.com> writes:
>
> > +/* a variable length array (i.e. _type *_field) but we know the
> > + * length
> > + */
> > +#define VMSTATE_STRUCT_VARRAY_POINTER_KNOWN(_field, _state, _num,
> > _version, _vmsd, _type) { \
> [...]
>
> Thinking about it some more, wouldn't VMSTATE_STRUCT_ARRAY_POINTER be a
> better name? Like with VMSTATE_ARRAY, the size of the array is known at
> compile-time. It's just that you need to dereference it first, hence
> ..._POINTER. There's nothing variable about it at all.
t's all a bit confusing; but the only pattern I'd figured out was that the
things after the 'VARRAY_' part tended to be talking about the length of
the array rather than the contents.
> But keep in mind I don't understand the current naming scheme in the
> first place, e.g. VMSTATE_ARRAY_INT32_UNSAFE vs. VMSTATE_VARRAY_INT32,
> with both of them specifying VMS_VARRAY_INT32...
No, I don't really either; one for Juan or Amit to suggest if they
prefer one or the other.
Dave
>
> Sascha
> --
> Softwareentwicklung Sascha Silbe, Niederhofenstraße 5/1, 71229 Leonberg
> https://se-silbe.de/
> USt-IdNr. DE281696641
>
--
Dr. David Alan Gilbert / dgilb...@redhat.com / Manchester, UK
