Update (Move bugs to potential).

git-svn-id: svn://scm.gforge.inria.fr/svn/mpfr/trunk@3516 280ebfd0-de03-0410-8827-d642c229c3f4

1 files changed, 11 insertions, 15 deletions

diff --git a/BUGS b/BUGS
index fa57a8ae2..e9939e404 100644
--- a/BUGS
+++ b/BUGS
@@ -24,35 +24,31 @@ Probably many bugs.
 
 Known bugs:
 
-* The overflows/underflows are partially implemented in some functions.
-  For instance, mpfr_pow (z, x, y, rnd) fails for very small x and some
-  values of y.
 
-* The mpfr_set_ld function assumes that the long double type has an
-  exponent of at most 15 bits.
-
-* When MPFR is compiled with Visual C++ under MS Windows, some tests may
-  fail, because of the way special IEEE-754 double values are tested.
 
 Potential bugs:
 
-* Possible integer overflows on some machines. Some types are used
-  inconsistently. Possible bugs with huge precisions (> 2^30).
+* The overflows/underflows exceptions may be badly handled in some functions.
+
+* Possible integer overflows on some machines.
+
+* Possible bugs with huge precisions (> 2^30).
 
 * Possible bugs if the chosen exponent range does not allow to represent
-  some numbers such as 1 or 1/2.
+  the range [1/16, 16].
 
 * Possible infinite loop in some functions for particular cases: when
   the exact result is an exactly representable number or the middle of
-  consecutive two such numbers.
+  consecutive two such numbers (for example, erf).
+
+* The mpfr_set_ld function may be quite slow if the long double type has an
+  exponent of more than 15 bits.
 
-* mpfr_set_d may give wrong results on some architectures.
+* mpfr_set_d may give wrong results on some non-IEEE architectures.
 
 * Error analysis for some functions may be incorrect (out-of-date due
   to modifications in the code?).
 
-* mpfr_hypot may fail for x very large, y very small and a very large
-  target precision. Other functions may be affected by similar problems.
 
 Problems due to compiler bugs: