This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: bug in g77
- To: Toon Moene <toon at moene dot indiv dot nluug dot nl>
- Subject: Re: bug in g77
- From: Steven Castillo <scastill at nmsu dot edu>
- Date: Wed, 14 Feb 2001 17:19:59 -0700 (MST)
- cc: gcc-bugs at gcc dot gnu dot org
On Tue, 13 Feb 2001, Toon Moene wrote:
> Steven Castillo wrote:
>
> > I am trying to run some fortran codes requiring very large arrays (~ 1
> > gbyte). When trying to compile a fortran code with a declared single
> > precision floating point array having a length of more than 134217727
> > using the command:
>
> Yes, this is a well-known limitation of g77: It calculates all sizes of
> objects in bits, so on 32-bit targets, the size of a REAL array is
> limited to 2**32/32 - 1= 2**27 - 1 = 134217728 - 1 = 134217727 elements.
>
> Your choices:
>
> 1. Help us fix this.
>
> 2. Switch to a 64-bit target, like Alpha.
>
> [ Hint: The latter may be cheaper ]
>
>
Hello Toon:
Thank you for your prompt reply! I think that you pose an interesting
problem that we may want to take a look at. I am not sure I agree with
your second possible solution. Give me a hint at where in the source code
to look and we may give it a shot.
Regards,
--
Steve Castillo
Professor and Department Head
Klipsch School of Electrical and Computer Engineering
New Mexico State University
scastill@nmsu.edu
(505)646-3117