https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103901
Jakub Jelinek <jakub at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jakub at gcc dot gnu.org, | |jason at gcc dot gnu.org, | |ppalka at gcc dot gnu.org --- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> --- If it is valid, I think e.g. --- gcc/cp/parser.c.jj 2022-01-04 09:59:37.297641779 +0100 +++ gcc/cp/parser.c 2022-01-04 20:56:19.065070397 +0100 @@ -11046,6 +11046,9 @@ cp_parser_lambda_expression (cp_parser* bool save_in_consteval_if_p = in_consteval_if_p; in_consteval_if_p = false; + int saved_processing_template_parmlist = processing_template_parmlist; + processing_template_parmlist = 0; + /* By virtue of defining a local class, a lambda expression has access to the private variables of enclosing classes. */ @@ -11078,6 +11081,7 @@ cp_parser_lambda_expression (cp_parser* in_consteval_if_p = save_in_consteval_if_p; in_discarded_stmt = discarded; + processing_template_parmlist = saved_processing_template_parmlist; parser->num_template_parameter_lists = saved_num_template_parameter_lists; parser->in_statement = in_statement; should fix that. But as the stack-overflow link says, it is unclear. Also, we still reject even with the above patch decltype ([]{ struct A {} ; return 1; }) a; (guess that is PR101911).