Okay.
Changing
E = 0.889994084835052D0
to
E = 0.889994084835051D0
or
E = 0.889994084835053D0
breaks it (no wonder the PowerPC chip behaves differently)!
Tim
On Wed, 31 Aug 2005, Patrick Wallace wrote:
> On Wed, 31 Aug 2005, Tim Jenness wrote:
>
>> * On my linux system it does run fine (status=0) *unless* I remove
>> the D0 from variable E (then it fails with status=-5). So two
>> runs, one with and one without the D0 would be great.
>
> I'm not too keen on the "removing the D0" part; something more controlled
> is needed. The variable E is double precision; omitting the D0 makes
> the constant single precision. Some compilers handle this by spotting
> that the constant is really double, but others will create a real and cast
> it.
>
> Try setting E to different double precision values and see if you can
> reproduce the instability that way.
>
>
> Patrick Wallace
> ____________________________________________________________________________
> SSTD/HMNAO Internet: [log in to unmask]
> Rutherford Appleton Laboratory Tel: +44-1235-445372
> Chilton, Didcot, Fax: +44-1235-446362
> Oxon OX11 0QX, UK
> ____________________________________________________________________________
>
--
Tim Jenness
JAC software
http://www.jach.hawaii.edu/~timj
|