Re: [PATCH] std-gnu23: new module

On 2024-11-16 16:06, Bruno Haible wrote: In the config.log I still see a warning: conftest.c:74:21: warning: 'null_pointer' defined but not used [-Wunused-const-variable=] 74 | constexpr nullptr_t null_pointer = nullptr; | ^~~~ Maybe you want to silence t

Re: [PATCH] std-gnu23: new module

Paul Eggert wrote: > as you mentioned we should use the > --std=gnu23 option anyway. I installed the attached to do that. Thanks. Both patches work fine. Now the gcc option is found and activated: checking for /gnu-inst-gcc/15-20241113/bin/gcc -ftrapv option to enable C23 features... -std=gnu

Re: [PATCH] std-gnu23: new module

On 11/16/24 05:48, Bruno Haible wrote: This test for is not useful, because is part of the C library, not part of the compiler. As I recall, I drafted that by looking at what's required even for freestanding compilers. Although the platforms you mentioned aren't up to supporting even freest

Re: [PATCH] std-gnu23: new module

On 11/16/24 05:23, Bruno Haible wrote: What is this line meant to test? - Support for hex escapes in UTF-8 string literals? - Comparison of bool values against int values? The former. Thanks for reporting that; I installed the attached.From 69d4a320a06e158d2f1e5df678186f2e84a9f6d2 Mon Sep

Re: [PATCH] std-gnu23: new module

Paul Eggert wrote in : > +# Prefer GNU C23 to earlier versions. This module does not have the intended effect: With a gcc 15 snapshot, the 'configure' script of a gnulib testdir produces: checking for /gnu-inst-gcc/15-20241113/b

Re: [PATCH] std-gnu23: new module

Paul Eggert wrote in : > +void c23_noreturn (); > + > +bool use_u8 = !u8"\xFF" == u8'\''x'\''; When configuring a gnulib testdir with a gcc 15 snapshot, this line produces two warnings: conftest.c:38:25: warning: comparison of c