[MITgcm-devel] lab_sea ad test

[Jean-Michel Campin]

Hi,

I see 4 possible solutions, all using the locally build cyrius makedepend:
1) add a specific AD_OPTFILE in lab_sea/code_ad + a genmake_local in
 lab_sea/build (=> it's specific to lab_sea AD test only)
2) only for lab_sea, but for all build, using the genmake_local; 
 this is what is beeing currently tested. Note that when makedepend 
 is specified in the optfile, the genmake_local has no effect.
 I made this change yesterday to see how it works on different platforms.
3) add an option to testreport to specify the makedepend to use.
 (-> would be used only for AD testing)
4) add the cyrius makedepend in the optfile (linux_ia32_g77).

A preference ?

Jean-Michel

On Thu, Feb 07, 2008 at 01:50:25AM -0500, Patrick Heimbach wrote:
> 
> It seems we were operating just under the MAXFILES limit,
> and are now above it with all the files that are used
> for this verif.
> I've seen this problem recently as well.
> -p.
> 
> Quoting Dimitris Menemenlis <menemenlis at sbcglobal.net>:
> 
> >Jean-Michel, do you know when this happened?
> >I don't think that it is related to recent pkg/seaice or pkg/exf
> >changes as I ran the lab_sea AD test before check-in.  D.
> >
> >>Hi,
> >>
> >>Looks like the AD test on faulks for lab_sea experiment is failling
> >>at the make depend level with this error message:
> >>>> makedepend: error:  out of space: increase MAXFILES
> >>>> make: *** [depend] Error 1
> >_______________________________________________
> >MITgcm-devel mailing list
> >MITgcm-devel at mitgcm.org
> >http://mitgcm.org/mailman/listinfo/mitgcm-devel
> >
> 
> 
> 
> --------------------------------------------------------
> Patrick Heimbach   Massachusetts Institute of Technology
> FON: +1/617/253-5259                  EAPS, Room 54-1518
> FAX: +1/617/253-4464             77 Massachusetts Avenue
> mailto:heimbach at mit.edu               Cambridge MA 02139
> http://www.mit.edu/~heimbach/                        USA
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel