Fwd: GNATS on Mac OS X

Here's a problem on PPC OS X, 10.4.6, GCC 4.0.1.

Perhaps someone with some autoconf skills can fix this.

cheers
peter

Begin forwarded message:

> From: Peter Gammie <[email protected]>
> Date: 12 May 2006 10:15:11 AM
> To: Toby Peterson <[email protected]>
> Subject: Re: GNATS on Mac OS X
>
> On 12/05/2006, at 10:02 AM, Toby Peterson wrote:
>
>> The error is caused by -Dconst=
>
> OK. Here's the test from gnats/configure.in:
>
> # Some Solaris C compilers have const problems without a special  
> ANSI option
> AC_CACHE_CHECK([for really working const],gnats_really_working_const,[
> AC_TRY_COMPILE([],
>                [const char *foo(const char *x) { return x; }
>                 int bar(const char *x) { return x == foo(x); }],
>                [gnats_really_working_const=yes],
>                [gnats_really_working_const=no])
> if test "$gnats_really_working_const" = no; then
>   CFLAGS="$CFLAGS -Dconst= "
> fi])
>
> and the result from gnats/config.log:
>
> configure:3993: checking for really working const
> configure:4016: gcc -c -I/sw/include -I/opt/local/include -I/usr/ 
> X11R6/include -I/sw/include -I/opt/local/include -I/usr/X11R6/ 
> include conftest.c >&5
> conftest.c: In function 'main':
> conftest.c:26: error: nested functions are not supported on MacOSX
> conftest.c:27: error: nested functions are not supported on MacOSX
> configure:4022: $? = 1
> configure: failed program was:
> | /* confdefs.h.  */
> |
> | #define PACKAGE_NAME "gnats"
> | #define PACKAGE_TARNAME "gnats"
> | #define PACKAGE_VERSION "4.1.0"
> | #define PACKAGE_STRING "gnats 4.1.0"
> | #define PACKAGE_BUGREPORT "[email protected]"
> | #define YYTEXT_POINTER 1
> | #define STDC_HEADERS 1
> | #define HAVE_SYS_TYPES_H 1
> | #define HAVE_SYS_STAT_H 1
> | #define HAVE_STDLIB_H 1
> | #define HAVE_STRING_H 1
> | #define HAVE_MEMORY_H 1
> | #define HAVE_STRINGS_H 1
> | #define HAVE_INTTYPES_H 1
> | #define HAVE_STDINT_H 1
> | #define HAVE_UNISTD_H 1
> | #define HAVE_ALLOCA_H 1
> | #define HAVE_ALLOCA 1
> | /* end confdefs.h.  */
> |
> | int
> | main ()
> | {
> | const char *foo(const char *x) { return x; }
> |                 int bar(const char *x) { return x == foo(x); }
> |   ;
> |   return 0;
> | }
> configure:4049: result: no
>
> Hmm... that seems completely spurious. It's unnecessary to nest the  
> functions like that, surely. That problem is fixed by just nuking  
> the test. Things look OK now.