Jason Merrill <ja...@redhat.com> writes:
> I'm inclined to change the C++ FE to pass NULL_TREE instead until such
> time as someone cares.

The sh backend will at least not choke on that ;-)

Reply via email to