[RFC] Use HAVE_LONGLONG instead of __GNUC__ to check for long long support.

M.M. Kettenis m.m.kettenis@alumnus.utwente.nl
Tue Apr 19 07:58:00 GMT 2005


Manoj Iyer <manjo@austin.ibm.com> wrote:

>
> Daniel,
>
> Used autoheader to generate the config.in file, but it does not look
> pretty, also, removed the comments that you did not like.
>
> 2005-04-18  Manoj Iyer  <manjo@austin.ibm.com>
>         * configure.ac: Added check for long long.
>         * config.in: Regenerated.
>         * configure: Regenerated.
>         * words.h: Modified logic to check for HAVE_LONGLONG instead of
>         __GNUC__, added config.h header file.

Please use AC_CHECK_TYPES, it should take care of defining
HAVE_LONG_LONG (note the extra underscore) and the comment all by
itself.

> --- src/sim/ppc/configure.ac	2005-01-12 10:23:21.000000000 -0600
> +++ new/src/sim/ppc/configure.ac	2005-04-18 18:34:10.000000000 -0500
> @@ -681,6 +681,9 @@ if test $ac_cv_struct_statfs = yes; then
>  	    [Define if struct statfs is defined in <sys/mount.h>])
>  fi
>
> +AC_CHECK_TYPE(long long,
> +   AC_DEFINE(HAVE_LONGLONG, 1, [Define if compiler supports long long.]),)
> +
>  dnl Figure out if /dev/zero exists or not
>  sim_devzero=""
>  AC_MSG_CHECKING(for /dev/zero)



More information about the Gdb-patches mailing list