[MITgcm-support] MITgcm_contrib

Mon Apr 21 12:09:51 EDT 2008

Elja,
I second all of Martin's suggestions in his latest mail.
-Patrick


On Apr 21, 2008, at 11:56 AM, Martin Losch wrote:
> I did check, but maybe your code is too old (calendardumps was  
> moved at some point). Try a fresh version and see if it works.
>
> I have no clue about the CVS, maybe you are trying to overwrite an  
> old checkout? I used CVS today without any problems.
>
> Start in a new directory.
> Martin
>
> On 21 Apr 2008, at 17:41, Elja Huibregtse wrote:
>> Hi Martin,
>>
>> Thanks for your reply! Unfortunately, it still doesn't work.
>>
>> 1) Trying: cvs co Mitgcm, leads again to:
>> cvs checkout: existing repository /u/gcmpack/CVSROOT/Emptydir does  
>> not match /u/gcmpack/MITgcm
>> cvs checkout: ignoring module MITgcm
>> connect to address 128.31.2.31: Connection timed out
>> Trying krb4 rsh...
>> connect to address 128.31.2.31: Connection timed out
>> trying normal rsh (/usr/bin/rsh)
>> forge.csail.mit.edu: Connection timed out
>> cvs [checkout aborted]: end of file from server (consult above  
>> messages if any)
>>
>>
>> 2) I did the make depend step and indeed a link to cal.h was  
>> created. However, I just found out that (at  least in my  
>> version...) calendarDumps is not created in cal.h but in model/inc/ 
>> PARAMS.h (PARM_L). Furthermore, it turns out that several  
>> variables are not declared (hdiag, enum_som_limiter). However,  
>> declaring them, leads to new errors.
>>
>> I really don't understand what is going on, although I think I am  
>> using wrong code-files and that CVS doesn't work properly.
>> Do you have some new suggestions?
>>
>> Thanks, I really appreciate your help!
>>
>> Elja
>>
>>
>>
>> -----Original Message-----
>> From: mitgcm-support-bounces at mitgcm.org on behalf of Martin Losch
>> Sent: Mon 4/21/2008 2:57 PM
>> To: mitgcm-support at mitgcm.org
>> Subject: Re: [MITgcm-support] MITgcm_contrib
>>
>> Hi Elja,
>>
>> the problem may not be connected at all.
>>
>> The first one I can't quite figure out, but it probabbly goes away,
>> when you checkout "everything" (cvs co MITgcm), and then only keep
>> global_ocean.cs32x15 and tutorial_held_suarez_cs/input in MITgcm/
>> verification.
>>
>> The second one seems to be a problem with finding files. Did you do
>> the "make depend" step? "calendardumps" is declared in pkg/cal/cal.h
>> and a link to this file (and many other *.F and *.h files) should
>> appear in your build directory after the "make depend" step.
>>
>> Martin
>>
>> On 21 Apr 2008, at 13:45, Elja Huibregtse wrote:
>> > Hi all,
>> >
>> > I am trying to set up the cube-sphere integration with sea-ice on
>> > the cs32.grid by applying the steps described in:
>> > http://mitgcm.org/cgi-bin/viewcvs.cgi/MITgcm_contrib/high_res_cube/
>> > README.cs32
>> >
>> > While applying "cvs co MITgcm_code", "cvs co MITgcm/verification/
>> > global_ocean.cs32x15" and "cvs co MITgcm/verification/
>> > tutorial_held_suarez_cs/input", I get the error message:
>> >
>> > -----------------
>> > cvs checkout: existing repository /u/gcmpack/CVSROOT/Emptydir does
>> > not match /u/gcmpack/MITgcm/utils
>> > cvs checkout: ignoring module MITgcm/utils
>> > connect to address 128.31.2.31: Connection timed out
>> > Trying krb4 rsh...
>> > connect to address 128.31.2.31: Connection timed out
>> > trying normal rsh (/usr/bin/rsh)
>> > forge.csail.mit.edu: Connection timed out
>> > cvs [checkout aborted]: end of file from server (consult above
>> > messages if any)
>> > ------------------
>> >
>> > Although it seems like something is really going wrong, I am able
>> > to continue to build up the 1-cpu cube sphere. Compiling the
>> > program ("make -j") fails, however:
>> >
>> > ------------------
>> > pgf90 -r8 -Mextend -O3 -LNO:fusion=2 -c cal_time2dump.f
>> > pgf90 -r8 -Mextend -O3 -LNO:fusion=2 -c diagnostics_addtolist.f
>> > pgf90 -r8 -Mextend -O3 -LNO:fusion=2 -c diagnostics_init_varia.f
>> > pgf90 -r8 -Mextend -O3 -LNO:fusion=2 -c diagnostics_interp_p2p.f
>> > PGF90-S-0038-Symbol, calendardumps, has not been explicitly
>> > declared (cal_time2dump.f)
>> >   0 inform,   0 warnings,   1 severes, 0 fatal for cal_time2dump
>> > make: *** [cal_time2dump.o] Error 2
>> > make: *** Waiting for unfinished jobs....
>> > PGF90-S-0072-Assignment operation illegal to external procedure
>> > hdiag (diagnostics_addtolist.f: 922)
>> > PGF90-S-0038-Symbol, hdiag, has not been explicitly declared
>> > (diagnostics_addtolist.f)
>> >   0 inform,   0 warnings,   2 severes, 0 fatal for
>> > diagnostics_addtolist
>> > make: *** [diagnostics_addtolist.o] Error 2
>> > ------------------
>> >
>> > Do I have to declare the symbols myself? Or do the versions of
>> > MITgcm (I am using checkpoint58) and the cube-sphere part not
>> > match? How can I fix it, any ideas?
>> >
>> > Thanks for your help!
>> >
>> > Elja
>> >
>> >
>> > _______________________________________________
>> > MITgcm-support mailing list
>> > MITgcm-support at mitgcm.org
>> > http://mitgcm.org/mailman/listinfo/mitgcm-support
>>
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>>
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support

---
Patrick Heimbach | heimbach at mit.edu | http://www.mit.edu/~heimbach
MIT | EAPS 54-1518 | 77 Massachusetts Ave | Cambridge MA 02139 USA
FON +1-617-253-5259 | FAX +1-617-253-4464 | SKYPE patrick.heimbach