This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: IMA corner case with forward declarations of statics
Dale Johannesen <dalej@apple.com> writes:
> I really don't think IMA mode ought to affect semantics. The
> result of building and running a program in IMA mode should be the
> same as with separate compilation, but possibly better optimized, IMO.
I agree in principle, but I do not think it is feasible in practice
and in the present state of the implementation. In *this* case,
realize that the function you patched (merge_translation_unit_decls)
no longer exists and the architecture is completely different (it
operates on a merge-as-you-go model). We could revert to the old
model but that will be substantially more work. Nor do I think this
is the only case where IMA will have to be pickier than normal
compilation.
(Mind, I'm just fine with making this a hard error in both modes.)
zw