On Tue, 18 Mar 2025, Martin Uecker wrote: > > I think different alignment should be considered different types, so an > > error. > > We currently allow different alignment to be added in typedefs > as in the new typedef-redecl3.c test:
I don't think we should. Typedef reclarations must have the "same type", which is supposed to avoid having any different information that would need merging. And whatever is unclear about the meaning of "same type" in some cases in the C standard, it seems clear that this case, though outside the scope of the standard, is definitely a case of different types because alignment information would need to be merged. (Unless the attributes here are part of the typedef that's not part of the associated type, but I don't think that's the case for alignment attributes - though it might be for the "deprecated" attribute, for example.) -- Joseph S. Myers josmy...@redhat.com
