Re: NEC-LIST:NEC 4.1 & g77

> Hello, all. Just thought I'd bring everyone up-to-date on my efforts
> with RH Linux, g77, and NEC 4.1. I never did hear back from Jurgen Von
> Hagen but this is the command that I originally used to compile NEC 4.1

sorry I was bussy getting a new job

> I used the g77 compiler (p/o GCC-2.95) that comes with the Red Hat Linux
> 7.1 distribution CDs to compile NEC 4.1D. The g77 compiler options were
> those used by Jurgen Von Hagen in compiling NEC 2D (re ACES Newsletter
> Nov '98):
>
> g77 -O3 -fno-emulate-complex -funroll-loops nec41d.f
>
> While the compile process issued a couple of warnings about same name
> for global and intrinsic and array size, the compile completed without
> error.

g77 is quite picky about some things that get unnoticed by other
compilers. The same name for global and intrinsic buisness is
about some functions that are defined inside nec and also defined
by g77. second (the timing routine) is one of these routines.

> I am using Red Hat 7.1 on a Panasonic Toughbook CF-27. Now, noticing a
> recent posting from Ray Anderson I tried compiling NEC 4.1 with
>
> g77 -O3 -fno-automatic nec41d.f

no-automatic is quite often needed as many fortran programmers
tend to forget that f77 variables are automatic - this
can lead to problems in some routines.

> and this worked. I'm not sure why Jurgen was using the
> -fno-emulate-complex instruction but that (not -funroll-loops) appears
> to be at the heart of the problem I was encountering.

no-emulate-complex does not emulate complex computations. g77
is based on the c backend which has a faulty complex
computation engine in some versions. Hence, g77 emulates
the complex computations and does not use the faulty c
routines. This happens only in some compilers, not others.
At the time I was compiling nec with g77 I had an engine that
was working ok - using the -fno-emulate-complex used faster
complex computations with no errors. So it seems to me
that you just had bad luck.

> The only other issue I have to resolve is that I if I compile NEC 4.1
> for more than 3500 segments (MAXSEG=3500, MAXMAT=3500) the NEC
> executable will cause a segmentation fault/core dump. Any thoughts in
> this area would certainly be welcome. Sincerely,

I have successfully used nec 4.1 (no longer, no longer in the US
so I can't help here as I cannot legally obtain use and work
with NEC 4.1) with as much as 6500 unknowns. No problem
has occured. Have you enough memory?

good to hear that you got it running

cheers
juergen

-- 
Institut fuer Hoechstfrequenztechnik und Elektronik - Uni Karlsruhe
Tel: +49 721 608 76 76, Fax: +49 721 69 18 65
-- 
The NEC-List mailing list <nec-list_at_gweep.ca>
http://www.gweep.ca/mailman/listinfo.cgi/nec-list