This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [C++ Patch/RFC PR c++/57887 (and dups)


On 11/14/2013 07:59 AM, Paolo Carlini wrote:
/* Make sure that any template parameters are in scope. */
maybe_begin_member_template_processing (decl);

Hmm, it seems that we use this for non-template member functions of template classes as well; I wonder when that happened. I guess we should update the comment to reflect reality. And perhaps the name of the function, though I lean slightly toward leaving that alone.

it seems clear to me that something very similar must be done for NSDMIs
too. Thus I prepared the below, which uses CLASSTYPE_TEMPLATE_INFO to
adapt the existing code in maybe_begin_member_template_processing. It
tests fine. How does it look?

I don't think we need a new parameter; just pass the FIELD_DECL into maybe_end_member_template_processing and adjust it appropriately.

Also, call m_e_m_t_p from cp_parser_late_parsing_nsdmi rather than cp_parser_class_specifier_1.

Jason



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]