[src/hypot.c] Added a comment on the FIXME.

git-svn-id: svn://scm.gforge.inria.fr/svn/mpfr/trunk@8586 280ebfd0-de03-0410-8827-d642c229c3f4

1 files changed, 8 insertions, 1 deletions

diff --git a/src/hypot.c b/src/hypot.c
index 1753541e1..0b75ed4d2 100644
--- a/src/hypot.c
+++ b/src/hypot.c
@@ -146,7 +146,14 @@ mpfr_hypot (mpfr_ptr z, mpfr_srcptr x, mpfr_srcptr y, mpfr_rnd_t rnd_mode)
      FIXME: Friedland in "Algorithm 312: Absolute Value and Square Root of a
      Complex Number" (Communications of the ACM, 1967) avoids overflow by
      computing |x|*sqrt(1+(y/x)^2) if |x| >= |y|, and |y|*sqrt(1+(x/y)^2)
-     otherwise. */
+     otherwise.
+     [VL] This trick (which is a scaling by a non-power of 2, thus doesn't
+     really bring new behavior w.r.t. overflow/underflow exceptions) may be
+     useful for hardware floating-point formats because a whole power-of-2
+     scaling code is likely to take more time than the additional division,
+     but in the context of multiple-precision, I doubt that it is a good
+     idea. Ideally scaling by a power of 2 could be done in a constant time,
+     e.g. with MPFR_ALIAS; but one needs to be very careful... */
   sh = mpfr_get_emax () / 2 - Ex - 1;
 
   MPFR_ZIV_INIT (loop, Nt);