[MITgcm-support] can't compile checkpoint 63m on Mac OSX (Jean-Michel Campin)

[Madeline Miller]

Hi Jean-Michel, Martin, Dimitris and Jody,

I should have updated; yes, the issue was solved and there is no problem
with the code, just with my memory.

Basically, I had forgotten that I was no longer using the same compilers
that Martin is.

Several months ago I accidentally wrote over the compilers when trying to
update them and then could neither get the update to work nor recover them.
In the end I had to revert to an ancient set of compilers (4.2.x) that
require
CFLAGS='-arch i386'. Previously I hadn't needed the flag because I had
newer (4.6.x) compilers.

Thanks for all the helpful suggestions.

Madeline


>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Tue, 29 May 2012 12:14:34 -0400
> From: Jean-Michel Campin <jmc at ocean.mit.edu>
> To: mitgcm-support at mitgcm.org
> Subject: Re: [MITgcm-support] can't compile checkpoint 63m on Mac OSX
> Message-ID: <20120529161434.GA6219 at ocean.mit.edu>
> Content-Type: text/plain; charset=us-ascii
>
> Hi Madeline,
>
> Did this issue got solved ?
> I am surprised that the code (c63m) does not compile anymore
> but the older version (c63j) was working fine (same platform, same
> compiler).
> Did you check carefully all the customized/config-specific files
> you have in your code dir ?
>
> Cheers,
> Jean-Michel
>
> On Fri, May 18, 2012 at 12:38:30PM +0200, Martin Losch wrote:
> > Hi Madeline,
> >
> > this is what I do in such a case.
> > 1. make sure that I really started from scratch, i.e.
> > make makefile && make CLEAN && make depend && make
> > 2. check out different versions of the MITgcm starting from 63j ( in
> bash:
> > for version in ( 63j 63k 63l 63m ); do
> >  cvs co -r checkpoint$version MITgcm_code && mv MITgcm ckpt$version
> > done
> > )
> > and find out when the problem starts.
> >
> > Also did you try to uncomment
> > CFLAGS='-arch i386'
> > ?
> >
> > Martin
> >
> > On May 18, 2012, at 2:59 AM, Madeline Miller wrote:
> >
> > > Dear List,
> > >
> > > I updated from checkpoint 63j to 63m and can no longer compile on my
> laptop (Mac OSX, Snow Leopard, ia386). I am using the default option file
> > > darwin_ia32_gfortran, but I also tried to compile with the c63j
> version of darwin_ia32_gfortran, to no avail. I am pasting the error below.
> Any ideas for how to fix or work around this?
> > >
> > > ld: warning: in gsl_ieee_env.o, file was built for unsupported file
> format which is not the architecture being linked (i386)
> > > ld: warning: in ptwrapper.o, file was built for unsupported file
> format which is not the architecture being linked (i386)
> > > ld: warning: in setrlstk.o, file was built for unsupported file format
> which is not the architecture being linked (i386)
> > > ld: warning: in sigreg.o, file was built for unsupported file format
> which is not the architecture being linked (i386)
> > > ld: warning: in tim.o, file was built for unsupported file format
> which is not the architecture being linked (i386)
> > > ld: warning: in timer_stats.o, file was built for unsupported file
> format which is not the architecture being linked (i386)
> > > Undefined symbols:
> > >   "_cusertime_", referenced from:
> > >       _timer_get_time__ in timers.o
> > >   "_timenow_", referenced from:
> > >       _timer_get_time__ in timers.o
> > >   "_csystemtime_", referenced from:
> > >       _timer_get_time__ in timers.o
> > > ld: symbol(s) not found
> > > collect2: ld returned 1 exit status
> > > make: *** [mitgcmuv] Error 1
> > >
> > > Best,
> > >
> > > Madeline
> > >
> > >
> > >
> > > _______________________________________________
> > > MITgcm-support mailing list
> > > MITgcm-support at mitgcm.org
> > > http://mitgcm.org/mailman/listinfo/mitgcm-support
> >
> > Martin Losch
> > Martin.Losch at awi.de
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20120530/8eb4c133/attachment.htm>