[mesa-users] Problems Installing

[Richard Townsend]

On Mar 6, 2013, at 3:57 PM, Aaron Dotter wrote:

> Hi all,
> 
> Guido and I spent some time on this issue (latest MESA with 32-bit Ubuntu) yesterday.  I think we have figured out what's going on.
> 
> The problem lies with allocating arrays, brought about by making the EOS tables 1-d arrays in the latest release.  After a while it just runs out of space.  It doesn't appear to happen in the eos test case because that doesn't try to load ALL of the tables.  star does.
> 
> In our case, setting 'ulimit -s unlimited' solved the EOS table problem but caused all sorts of problems in compiling and testing net.  So that's a no-go.  We reverted back to 4723 and all was well.
> 

Very interesting. I wonder if we're just running out of addressable memory. What exactly are the problems with net -- can you send some example output? 

> I guess the lesson is that using the latest MESA on 32-bit linux systems is going to be a problem under the new 1-d array regime.  I would be very interested to learn if there is a fix for this at the OS level, as in setting ulimit in some way that doesn't create new problems while solving old ones.
> 
> Aaron
> 
> 
> 
> On Thu, Mar 7, 2013 at 8:00 AM, Adam Lewis <adam.lws at gmail.com> wrote:
> Hello again,
> 
> I reinstalled with 64-bit Ubuntu 12.04 and the problem resolved. The other major difference is that no instances of e.g. gfortran exist on my system in locations besides that of the SDK.
> 
> -Adam
> 
> 
> On Tue, Mar 5, 2013 at 5:15 PM, Richard Townsend <townsend at astro.wisc.edu> wrote:
> Hi Adam --
> 
> You shouldn't overwite test_output -- that's not the problem, the problem is with tmp.txt.
> 
> You need to find out what's happening when tmp.txt is being created. I suggest you look through the relevant scripts and find out what errors are occurring.
> 
> cheers,
> 
> Rich
> 
> 
> On Mar 5, 2013, at 3:48 PM, Adam Lewis wrote:
> 
> > Hello,
> >
> > I am attempting to install mesa/star (v. 4798) on a computer running Ubuntu 12.04 32bit.
> >
> > I am using the SDK and have all the listed packages.
> >
> > When installing everything seems to work until the star package is built, at which point I receive the following output:
> > ***
> > /home/adam/work/mesa/star
> > building star package.
> >
> > make: `libstar.a' is up to date.
> > make: `libstar.a' is up to date.
> > gfortran -fno-range-check  -fopenmp  -o ../star \
> >     run_star_extras.o run_star.o  run.o -L../../make -lstar -L../../../lib -ldiffusion -lionization -latm -lcolors -lmlt -lnse -lnet -leos -lkap -lscreen -lrates -lreaclib -lweak -lneu -lchem -linterp_2d -linterp_1d -lnum -lutils -lalert -lmtx -lconst -lmesaklu `mesasdk_lapack_link` `mesasdk_blas_link` `mesasdk_pgplot_link` `mesasdk_se_link`
> >
> > first file is short
> >
> > /home/adam/work/mesa/star/test
> > TEST FAILED -- compare test_output to tmp.txt
> >
> >
> > /home/adam/work/mesa/star
> > ./build_and_test FAILED
> > ***
> >
> > cat /mesa/star/test/test_output yields
> > "stop because star_age >= max_age".
> >
> > I was able to bypass the test by copying tmp.txt into test_output. This permits compilation. I then modified the inlist to read from inlist_first_thing_to_try as described on the website and executed ./rn, generating the error message
> >
> > Load1_eosDT_Table ierr        5014           2           4  0.20000000000000001       0.20000000000000001
> >  Load1_eosDT_Table ierr        5014           1           1   0.0000000000000000        0.0000000000000000
> >  Load1_eosDT_Table ierr        5010           1           1   0.0000000000000000        0.0000000000000000
> >  Load1_eosDT_Table ierr        5010           1           1   0.0000000000000000        0.0000000000000000
> >  finish_load_model: failed in set_vars
> >  failed in finish_load_model
> >  star_load_zams ierr          -1
> > WARNING: allocate failed for eos tables
> >  do_load1_star ierr          -1
> > WARNING: allocate failed for eos tables
> > Tue Mar  5 16:46:52 EST 2013
> >
> > Any help would be greatly appreciated.
> >
> > -Adam
> >
> >
> >
> >
> >
> >
> >
> > ------------------------------------------------------------------------------
> > Symantec Endpoint Protection 12 positioned as A LEADER in The Forrester
> > Wave(TM): Endpoint Security, Q1 2013 and "remains a good choice" in the
> > endpoint security space. For insight on selecting the right partner to
> > tackle endpoint security challenges, access the full report.
> > http://p.sf.net/sfu/symantec-dev2dev_______________________________________________
> > mesa-users mailing list
> > mesa-users at lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/mesa-users
> 
> 
> 
> ------------------------------------------------------------------------------
> Symantec Endpoint Protection 12 positioned as A LEADER in The Forrester
> Wave(TM): Endpoint Security, Q1 2013 and "remains a good choice" in the
> endpoint security space. For insight on selecting the right partner to
> tackle endpoint security challenges, access the full report.
> http://p.sf.net/sfu/symantec-dev2dev
> _______________________________________________
> mesa-users mailing list
> mesa-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/mesa-users
> 
> 
> ------------------------------------------------------------------------------
> Symantec Endpoint Protection 12 positioned as A LEADER in The Forrester  
> Wave(TM): Endpoint Security, Q1 2013 and "remains a good choice" in the  
> endpoint security space. For insight on selecting the right partner to 
> tackle endpoint security challenges, access the full report. 
> http://p.sf.net/sfu/symantec-dev2dev_______________________________________________
> mesa-users mailing list
> mesa-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/mesa-users