Jakub,
The following patch implements P0386R1 - NSDMIs for bit-fields.
While working on that, I've discovered our parser mishandles attributes
on bitfields, already C++11 says:
identifier[opt] attribute-specifier-seq[opt] : constant-expression
in the grammar, but we actually parsed
identifier[opt] : constant-expression attribute-specifier-seq[opt]
I'm sorry for my tardiness. It think the patch would be better broken
apart:
1) fix the parsing bug you found and move to (ab)using
DECL_BIT_FIELD_REPRESENTATIVE
2) the new c++2a feature
Is that feasible?
WRT your questions
1, I think a default strict arg on cp_parser_constant_expression is
the way to go. That makes it clear that its default behaviour is
'sloppy', saving someone the detective work you've done.
2. I'm all for a pedwarn, Jason is often more conservative than me
though.
3. D_B_F_R is quite probably fine -- you'll know better than me,
having poked at it
4. They are not valid on unamed bitfields. You may have spotted an
inconsistency in the spec.
nathan
--
Nathan Sidwell
