[Ecls-list] BUG: floating point underflow in TAN on single-float

[Juan Jose Garcia-Ripoll]

On Sat, May 17, 2008 at 7:36 PM, Eric Marsden <eric.marsden at free.fr> wrote:
>> (tan (coerce (/ pi 4) 'single-float))
> #<a FLOATING-POINT-UNDERFLOW>
>
> ECL 0.9j (CVS 2008-05-13 17:26) on Linux/AMD64.

I know. This is a bug in the GNU C library which I reported like about
one year and a half. Assumed they would have solved it by now. It does
not happen on other platforms or architectures (See below) I will
write a test for the configuration process to detect it.

Juanjo

ECL is free software, and you are welcome to redistribute it
under certain conditions; see file 'Copyright' for details.
Type :h for Help.  Top level.
> (tan (coerce (/ pi 4) 'single-float))
1.0
> ^D
$ uname -a
Darwin benasque2 8.11.1 Darwin Kernel Version 8.11.1: Wed Oct 10
18:23:28 PDT 2007; root:xnu-792.25.20~1/RELEASE_I386 i386 i386
MacBook2,1 Darwin

> (tan (coerce (/ pi 4) 'single-float))
1.0
>
[jjgarcia at quantum jjgarcia]$ uname -a
Linux quantum.fis.ucm.es 2.4.18-14 #1 Wed Sep 4 13:35:50 EDT 2002 i686
i686 i386 GNU/Linux


Juanjo

-- 
Facultad de Fisicas, Universidad Complutense,
Ciudad Universitaria s/n Madrid 28040 (Spain)
http://juanjose.garciaripoll.googlepages.com