Re: [PATCH 0/3] Use gnulib c-ctype module, not <ctype.h>

[Simon Marchi <simark@simark.ca>]

On 8/6/25 9:13 AM, Tom Tromey wrote:
> This series changes gdb to use the gnulib c-ctype module rather than
> <ctype.h> or "safe-ctype.h".
> 
> The rationale is mostly in patch #3.  I do wonder if it would be
> better to just add the casts everywhere.
> 
> Regression tested on x86-64 Fedora 41.  This doesn't cover
> solib-rocm.c though.
> 
> Signed-off-by: Tom Tromey <tromey@adacore.com>

My understanding is that poisoning using macros doesn't work, because it
interferes with identifiers of the same name in other headers.  But
would it be possible to do it using "#pragma GCC poison"?  Like:

  #pragma GCC poison isspace

If I try to use it, I get:

      CXX    cp-support.o
    /home/smarchi/src/binutils-gdb/gdb/cp-support.c:1137:22: error: attempt to use poisoned ‘isspace’
     1137 |               while (isspace(name[index]))
          |                      ^
    In file included from /home/smarchi/src/binutils-gdb/gdb/defs.h:26,
                     from <command-line>:
    /home/smarchi/src/binutils-gdb/gdb/../gdbsupport/common-defs.h:231:20: note: poisoned here
      231 | #pragma GCC poison isspace
          |                    ^~~~~~~

This pragma can obviously be gated by an appropriate ifdef, if not all
compilers support it.

I was wondering if clang-tidy would have a check for this.  The answer
is no, but there is a bug open about it:

  https://github.com/llvm/llvm-project/issues/63280

Simon