[MITgcm-support] Building adjoint with OpenAD and gfortran

[Chris Wilson]

Dear Chris, Patrick and Utke,

Thanks for your help.  From Utke's comments, I guess that I've been (perhaps blindly) following the instructions to build the OpenAD example (global 4x4, tracer outgassing sensitivity) given in verification/OpenAD/README without relating this to other instructions for compilation of the particular AD tool as given in section 5.2 of the manual.  

I'm still a little unclear on what I need to do beyond the README instructions.  Should I try to build OpenAD using Open64 first?

I'd be grateful of any more detailed pointers you could give me please.  For now, I'll hold off taking up your kind offer of a test account on your Linux server, Chris.  

Apologies for my naivety.  I've included my attempts at compilation of the OpenAD example in some attachments.
-- 
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: notes.txt
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20100506/141a603f/attachment.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: darwin_ia32_gfortran_cw
Type: application/octet-stream
Size: 1012 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20100506/141a603f/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: darwin_ia32_gfortran_cw_ffixed-form
Type: application/octet-stream
Size: 1038 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20100506/141a603f/attachment-0001.obj>
-------------- next part --------------


Chris

+++++++++++++++++++++++++++++++++++++
Dr Chris Wilson

NERC National Oceanography Centre,
Joseph Proudman Building, 6 Brownlow Street, 
Liverpool L3 5DA, United Kingdom

Telephone: +44(0)151 795 4806
Reception: 4800; Fax: 4801

www.pol.ac.uk/home/staff/?user=WilsChri
www.nerc.ac.uk
+++++++++++++++++++++++++++++++++++++

On 5 May 2010, at 19:06, Jean Utke wrote:

> Hi,
> 
> yes, I got the Mac and right now I am waiting for a replacement because
> it turns out the machine I got has a problem with the display.
> In the meantime I got an old mac book as a  loaner that has
> some gfortran of unclear origin installed.
> The biggest hurdle for the Mac compile is Open64 which has
> an ancient build system that was not designed to be truly portable
> and neither was the code.
> I fixed a bunch of system header includes that pop up as early  problems
> but Open64 needs a fortran compiler for its constant folding etc.
> and that is AFAICT a more iffy thing because there is no standard
> Fortran compiler for Mac OS and various gfortran binaries offered
> as add-ons don't necessarily work as expected.
> Chris - is there some gfortran binary distribution for Mac you guys agree on
> or is it expected that people compile their own from scratch?
> 
> This aside we are moving from Open64 to Rose as a front-end
> anyway and therefore I am not sure how much effort I should put into
> making  Open64 compile on a Mac.
> 
> Jean
> 
> On 05/05/2010 01:46 AM, Chris Hill wrote:
>> Hi Chris
>> 
>> Jean who leads OpenAD development recently got a Mac.
>> He may have tried this.
>> 
>> I find building OpenAD somewhat complicated (there are a lot of other
>> packages it uses). I use it on a Linux server machine where it has already
>> been installed for me (by Jean). If it helps we can get you an
>> account on that machine,
>> so you can do a little testing and we can see if we can
>> help you get going.  You would need somewhere to do long runs eventually
>> though. Let us know.
>> 
>> Chris Hill
>> 
>> On Tue, May 4, 2010 at 2:24 PM, Wilson, Christopher <cwi at pol.ac.uk> wrote:
>> 
>>> Hi,
>>> 
>>> Does anyone have experience building the example in c61 verification/OpenAD using gfortran under OSX?
>>> 
>>> If so, I'll follow this message with the details of my problem.
>>> 
>>> Muchas gracias!
>>> 
>>> Chris Wilson
>>> National Oceanography Centre, Liverpool, UK
>>> 
>>> --
>>> This message (and any attachments) is for the recipient only. NERC
>>> is subject to the Freedom of Information Act 2000 and the contents
>>> of this email and any reply you make may be disclosed by NERC unless
>>> it is exempt from release under the Act. Any material supplied to
>>> NERC may be stored in an electronic records management system.
>>> 
>>> 
>>> _______________________________________________
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>>> 
>>> 
>