https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91507
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Keywords| |wrong-debug
CC| |jsm28 at gcc dot gnu.org
Known to fail| |10.0, 9.2.1
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
And the C frontend confuses dwarf2out here by calling
dwarf2out_early_global_decl
twice with the same decl but with different unrelated TREE_TYPE of it.
There's of couse no way it can be expected to recover from that? I guess
the situation can only arise for decls at global scope? So we could
in theory try to lookup the DIE for the decls type after looking up a
specification DIE and emit a new DIE for it? Ideally we'd have marked
the earlier DIE as "incomplete" somehow.
