Subject: Re: lib/32861: frexp returns incorrect value for small floating-point
To: None <gnats-bugs@netbsd.org>
From: Matthias Drochner <M.Drochner@fz-juelich.de>
List: netbsd-bugs
Date: 02/21/2006 17:57:18
Francois-Xavier.Coudert@ens.fr said:
> I was just investigating a problem encountered in the gfortran
>  testsuite

OK, so I've built a gcc-4.0.2 and run its testsuite.
intrinsic_fraction_exponent.f90 and intrinsic_set_exponent.f90
did fail on an unchanged system.
With my patch in place, both appear to succeed. (At least they
don't abort() if I run them standalone; I'm not so comfortable
with dejagnu to know how to run single tests.)
If there are more tests in unreleased gcc versions which fail
due to problems with subnormal numbers please let me know.

thanks for your report
best regards
Matthias