On Tue, 10 Jan 2006 15:44:41 +0000 Keith MARSHALL <keith.marshall@xxxxxxxxx> wrote: KM> Hmm. This trivial configure.ac: KM> KM> AC_INIT KM> AC_CHECK_FUNCS([snprintf _snprintf]) KM> AC_CONFIG_HEADERS([config.h]) KM> AC_OUTPUT KM> KM> results in a configure script which says "yes" for both snprintf KM> and _snprintf, both when I run it under Cygwin, and under MSYS/MinGW, KM> and it #defines both HAVE_SNPRINTF and HAVE__SNPRINTF appropriately, KM> in both cases. Indeed. I think that the results I was getting may have been due to --std=xxx option in use or something like that then. In any case, I'd also like to be able to use MSVC cl.exe eventually (I think it should be ok at autoconf level although it promises to be much more interesting with libtool) so I definitely need the test for _snprintf as fallback anyhow. Thanks, VZ _______________________________________________ Autoconf mailing list Autoconf@xxxxxxx http://lists.gnu.org/mailman/listinfo/autoconf
