Great, +1 for you!
Your NERSC system picked gcc compiler; deal.II compiles just fine on your
system, however, you want to do something more complex. Now we want to
check PETSc on NERSC is compatible with deal.II on NERSC.
Let's try to get PETSc working now.
You have two ways to go from here:
(i) Check that PETSc was compiled with gcc, in which case follow the
details in the deal.II README. Everything should be working just fine. If
not, go to (ii).
(ii) If PETSc was compiled with a different compiler (icc, or whatever)
you will need to compile dealii with that same compiler. Try compiling
deal.II with that compiler and *without PETSc*. Let's check the compilers
are ok for deal.II.
You can use ./configure --help to figure out how to change which compiler
deal.II will choose. Make sure if PETSc was compiled with static (or
dynamic) deal.II is compiled with static (or dynamic).
Dan, I am very happy to walk you through this procedure (you seem to be
having alot more trouble than "normal" users). I also think it is a waste
of resources to continue on the deal.II mailing list until we have a
result or a problem we can not solve.
Let's continue this discussion off list (email me directly) and if we run
into trouble we can ask the experts here. If all goes ok, we could think
about updating the deal.II wiki to help other NERSC users.
Let me know how your tests go with the compilers.
I hope this helps.
Best,
Toby
-----
Toby D. Young
Assistant Professor
Institute of Fundamental Technological Problems
Polish Academy of Sciences
ul Adolfa Pawinskiego 5b
02-106 Warsaw
Poland
www: http://www.ippt.gov.pl/~tyoung
skype: stenografia
On Wed, 6 Jul 2011, Daniel Brauss wrote:
> Thanks for the help Toby and Michael.
>
> I went and did a make distclean
> I did a ./configure
> and I did a make all
>
> Everything seemed to compile with no problem. I ran the first 5 step
> programs. All the output looked good. I have attached the the
> Make.global_options. What do you guys suggest?
>
> Dan
>
>
>
> On Wed, Jul 6, 2011 at 7:28 AM, Toby D. Young <[email protected]> wrote:
>
> >
> > Dan,
> >
> > How about we take one step backward, and maybe then we can take two steps
> > forward.
> >
> > Let's try to debug this thing from the beginning. This should be alot
> > faster and alot more sensible than trying to poke invisable monkeys in the
> > dark with a black stick.
> >
> > To help you, *you* have to help us; and that means giving us logical,
> > thought out information about your problem.
> >
> > Suggestion:
> >
> > Try downloading (or svn checkout) deal.II one more time (or run make
> > distclean).
> >
> > Just run ./configure, make all, and let us know what happens. Do not even
> > think about linking to external libraries yet. If deal.II compiles, try
> > the examples. If something goes wrong, send your
> > common/Make_global_options and error messages to [email protected].
> >
> > Start fresh. Even if you did that already, do it again and convince me
> > that it works (or does not).
> >
> > Then we can start working on the BOOST/PETSc/Whatever library problems.
> >
> > Best,
> > Toby
> >
> >
> > -----
> >
> > Toby D. Young
> > Assistant Professor
> >
> > Institute of Fundamental Technological Problems
> > Polish Academy of Sciences
> > ul Adolfa Pawinskiego 5b
> > 02-106 Warsaw
> > Poland
> >
> > www: http://www.ippt.gov.pl/~tyoung
> > skype: stenografia
> >
> > On Wed, 6 Jul 2011, Daniel Brauss wrote:
> >
> >
>
_______________________________________________
dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
