FX via Gcc <gcc@gcc.gnu.org> writes:
> I’m trying to bootstrap a GCC 10 compiler on macOS with clang, and I am
> getting errors in stage 1, because there is C++11 code that is rejected by
> clang (because the bootstrap involves compiling stage 1 with -std=gnu++98,
> online on master (see top-level configure.ac). These errors are not seen, I
> believe, when GCC is the bootstrap compiler, because GCC will issue a warning
> instead of an error (as clang does).
>
> One place with such issue is in aarch64-builtins.c, which contains a C++11
> constructor. I can fix it with this:
>
> diff --git a/gcc/config/aarch64/aarch64-builtins.c
> b/gcc/config/aarch64/aarch64-builtins.c
> index cba596765..184e9095d 100644
> --- a/gcc/config/aarch64/aarch64-builtins.c
> +++ b/gcc/config/aarch64/aarch64-builtins.c
> @@ -1225,8 +1225,9 @@ aarch64_init_memtag_builtins (void)
> = aarch64_general_add_builtin ("__builtin_aarch64_memtag_"#N, \
> T, AARCH64_MEMTAG_BUILTIN_##F); \
> aarch64_memtag_builtin_data[AARCH64_MEMTAG_BUILTIN_##F - \
> - AARCH64_MEMTAG_BUILTIN_START - 1] = \
> - {T, CODE_FOR_##I};
> + AARCH64_MEMTAG_BUILTIN_START - 1].ftype = T; \
> + aarch64_memtag_builtin_data[AARCH64_MEMTAG_BUILTIN_##F - \
> + AARCH64_MEMTAG_BUILTIN_START - 1].icode =
> CODE_FOR_##I;
>
> fntype = build_function_type_list (ptr_type_node, ptr_type_node,
> uint64_type_node, NULL);
>
> […stuff that Iain has already answered…]
>
> I would welcome:
>
> 1. confirmation that the C++11 code in aarch64-builtins.c is indeed a bug,
> and that a patch for it would be welcome
Yeah, it's definitely a bug, thanks for catching it. The patch above
is OK.
Thanks,
Richard
