I installed the following simple patch to work around the Autoconf problem you reported last month. I'm not sure the fancier solutions we discussed are worth the aggravation of developing and maintaining them. Thanks again for reporting the problem. 2004-08-19 Paul Eggert <eggert@xxxxxxxxxxx> * lib/autoconf/c.m4 (AC_LANG_BOOL_COMPILE_TRY(C), AC_C_LONG_DOUBLE): Undo 2004-06-04 change, as it didn't work with HP-UX 11.23 cc/aCC or Tru64 4.0 cc. Problem reported by Noah Misch in <http://lists.gnu.org/archive/html/autoconf/2004-07/msg00004.html>. Index: c.m4 =================================================================== RCS file: /cvsroot/autoconf/autoconf/lib/autoconf/c.m4,v retrieving revision 1.189 retrieving revision 1.190 diff -p -u -r1.189 -r1.190 --- c.m4 4 Jun 2004 19:39:54 -0000 1.189 +++ c.m4 20 Aug 2004 07:01:02 -0000 1.190 @@ -228,14 +228,13 @@ char (*f) () = $1; # AC_LANG_BOOL_COMPILE_TRY(C)(PROLOGUE, EXPRESSION) # ------------------------------------------------- -# The C standard does not require a diagnostic when we compute 1 / 0 -# at compile-time, but we don't know of any compiler that fails to -# diagnose this. Diagnostics are required for some other expressions -# (e.g., ((EXPRESSION) ? 0 : (0, 0)), INT_MAX + !!(EXPRESSION)), -# but they don't work as well in practice. +# Return a program that is valid if EXPRESSION is nonzero. +# EXPRESSION must be an integer constant expression. +# Be sure to use this array to avoid `unused' warnings, which are even +# errors with `-W error'. m4_define([AC_LANG_BOOL_COMPILE_TRY(C)], -[AC_LANG_PROGRAM([$1], [static int v = 1 / !!($2); -v = 0; +[AC_LANG_PROGRAM([$1], [static int test_array @<:@1 - 2 * !($2)@:>@; +test_array @<:@0@:>@ = 0 ])]) @@ -968,8 +967,6 @@ AC_DEFUN([AC_C_LONG_DOUBLE], + (DBL_MANT_DIG < LDBL_MANT_DIG) - (LDBL_MAX_EXP < DBL_MAX_EXP) - (LDBL_MANT_DIG < DBL_MANT_DIG))) - && (0 < ((DBL_MAX < LDBL_MAX) + (LDBL_EPSILON < DBL_EPSILON) - - (LDBL_MAX < DBL_MAX) - (DBL_EPSILON < LDBL_EPSILON))) && (int) LDBL_EPSILON == 0 ]])], ac_cv_c_long_double=yes, _______________________________________________ Autoconf mailing list Autoconf@xxxxxxx http://lists.gnu.org/mailman/listinfo/autoconf