| Message ID | a35ea4e7-79a1-15a5-9f71-d9aea316cf72@oracle.com |
|---|---|
| State | New |
| Headers | show |
OK. On Thu, Nov 17, 2016 at 3:25 PM, Paolo Carlini <paolo.carlini@oracle.com> wrote: > Hi, > > On 17/11/2016 18:14, Jason Merrill wrote: >> >> How about changing cp_parser_non_integral_constant_expression to issue >> a pedwarn instead of error for NIC_FLOAT? > > Indeed. And a relevant observation is that we don't pass NIC_FLOAT from > anywhere else in the parser, please correct me if I'm wrong. The below is > finishing testing, all good so far. > > Thanks, > Paolo. > > ////////////////
Index: cp/parser.c =================================================================== --- cp/parser.c (revision 242540) +++ cp/parser.c (working copy) @@ -3028,8 +3028,9 @@ cp_parser_non_integral_constant_expression (cp_par switch (thing) { case NIC_FLOAT: - error ("floating-point literal " - "cannot appear in a constant-expression"); + pedwarn (input_location, OPT_Wpedantic, + "ISO C++ forbids using a floating-point literal " + "in a constant-expression"); return true; case NIC_CAST: error ("a cast to a type other than an integral or " Index: testsuite/g++.dg/parse/pr55080.C =================================================================== --- testsuite/g++.dg/parse/pr55080.C (revision 0) +++ testsuite/g++.dg/parse/pr55080.C (working copy) @@ -0,0 +1,6 @@ +// PR c++/55080 +// { dg-options "-std=c++98 -pedantic" } + +class B { + static const int c = 3.1415926; // { dg-warning "constant-expression" } +};