Macros MPFR_EXP_INVALID (invalid exponent value) and MPFR_EXP_CHECK

added. Code update to use MPFR_GET_EXP and MPFR_SET_EXP instead of
MPFR_EXP to allow more bug detection related to special values.
Macros MPFR_SET_NAN, MPFR_SET_INF, MPFR_SET_ZERO and MPFR_INIT set
the exponent of the number to MPFR_EXP_INVALID if MPFR_EXP_CHECK
is defined. Compile with -DMPFR_EXP_CHECK and make check to see
the potential problems; currently, 40 of 76 tests fail.


git-svn-id: svn://scm.gforge.inria.fr/svn/mpfr/trunk@2301 280ebfd0-de03-0410-8827-d642c229c3f4

1 files changed, 2 insertions, 2 deletions

diff --git a/set_exp.c b/set_exp.c
index e3793403a..8fc628554 100644
--- a/set_exp.c
+++ b/set_exp.c
@@ -1,6 +1,6 @@
 /* mpfr_set_exp - set the exponent of a floating-point number
 
-Copyright 2002 Free Software Foundation.
+Copyright 2002, 2003 Free Software Foundation.
 
 This file is part of the MPFR Library.
 
@@ -29,7 +29,7 @@ mpfr_set_exp (mpfr_ptr x, mp_exp_t exponent)
 {
   if (exponent >= __gmpfr_emin && exponent <= __gmpfr_emax)
     {
-      MPFR_EXP(x) = exponent;
+      MPFR_EXP(x) = exponent; /* do not use MPFR_SET_EXP of course... */
       return 0;
     }
   else