Paul Eggert wrote: > Thanks, that sounds nicer. I did that. Thanks a lot!
> My earlier patch had been > assuming we couldn't use that C99 feature; I also documented that we're > assuming C99 empty macro args by installing the attached additional patch. Nice. My knowledge comes from the fact that GNU clisp still has an old configure test for the use of empty macro arguments, and I recently ran it on all available platforms and found that it passes everywhere. The comments in GNU clisp Some compilers (ie. cc under HP-UX) seem to interpret a macro call foo(arg1,...,argn,) as equivalent to foo(arg1,...,argn), which will yield an error. date back to 1998 or before; that's before C99 even existed. > The current code doesn't check for portability to strict C17 and > earlier, when compiling with recent GCC. Although it could be > complicated to do so I hope the complexity isn't necessary, as we can > check by compiling with stricter C17- compilers. I didn't try any compiler with strict C99 compliance. But there are no problems - with the usual gcc and clang versions used in the CI, - with gcc 3.3 as in Fedora 1, - with Sun C on Solaris 10, - with xlc on AIX 7.1. Bruno
