Subject: Re: porting ccm3 to linux
From: fei liu (feiliu@aa.eps.jhu.edu)
Date: Fri Sep 12 1997 - 07:36:02 MDT
Following is a tedious message about running ccm3 and the possibility
to port it to linux, if you are not interested in this issue, just
ignore it.
Fei Liu
Please send me a message if you are interested in this topic after
reading this email. Thanks.
----- Begin Included Message -----
>From erik@ra.cgd.ucar.edu Thu Sep 11 23:15:24 1997
Date: Thu, 11 Sep 1997 21:11:57 -0600 (MDT)
From: Erik Kluzek <erik@ra.cgd.ucar.edu>
X-Sender: erik@mercury
To: feiliu@aa.eps.jhu.edu
Subject: Re: Disk-space and memory for CCM
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Content-Length: 1917
> Actually, I am trying to port the ccm3 to a couple of systems, like
> linux, and OSF1 alpha machine. I also feel that's a memory space problem
> on aa.eps which is a sun machine. The problem I met there is the
> freemem and getmem.Fs which dynamically allocate and free memory for
> ccm3, and I can not compile pass these two programs. I have been doing
> well with around 100 more other Fortran source files in ccm3 package.
> I would be very glad if I can hear some suggestions on this issue.
Ok, that's an important "detail" to get past. It's also a pretty hard
thing to work out.
One suggestion I can think of is making sure you have the compiler
machine architecture setup for the machines you run on. We've got
defines setup for Cray, SGI, and RS6K. You'll need to get defines
setup for the new architetures. getmem and freemem especially are
completly compiler dependent, so you'll need to get them to do the
right thing...
>
> The problem stuck there on linux 2.0.0 with ccm3 package is the pointer
> syntax in getmem.F and freemem.F. If you would like to look at the problem
> a little bit further, I will send them to you in the next email.
>
Other then maybe giving suggestions I really can't dig into your problem.
But, this is an interesting problem, that others are interested in as well.
Create a message to pass onto the ccm-users list. I know of at least one
other person interested in getting ccm3 to work on Linux. So go ahead and
present the problem and the type of errors you are getting. The ccm-users
list is likely to bring out a lot more experience and interest in this
issue.
Anyway, it would be great to get the ccm on Linux. We'd be interested in
continuing to hear about your status...
Erik
Erik Kluzek, (CGD at NCAR)
National Center for Atmospheric Research
Boulder CO, (off) (303)497-1326 (fax) (303)497-1324
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
----- End Included Message -----
This archive was generated by hypermail 2b27 : Thu Jun 01 2000 - 09:11:30 MDT